Models, products and infrastructure, with source links and review status.
439 entries, including 48 marked for review. Latest publication 8 Nov 2026.
Showing 439 of 439 entries
Adobe
Adobe's agent for customer-experience orchestration, now runnable inside ChatGPT, Claude, Gemini, Microsoft Copilot and Databricks Genie.
Source record checked 8 Nov 2026
Cloudflare
Cloudflare adds a Personal Assistants analytics category and per-agent Allow/Block settings for agents acting on behalf of users.
Source record checked 8 Nov 2026
Fiddler AI
Agent evaluation, monitoring and explainability platform that CrowdStrike agreed to acquire on Nov 3, 2026, to sell as Falcon Observe.
Source record checked 8 Nov 2026
Google's video generation and editing model: native 15-second clips, multi-turn editing, $0.07 per second at 720p.
Source record checked 8 Nov 2026
Honeycomb
Goal-driven AI SRE agents for Honeycomb telemetry: a private Personal Agent and background Team Agents, in early access.
Source record checked 8 Nov 2026
Kong
Kong Konnect control plane for governing agentic AI: cost attribution, session observability, registry, token vault and agent identity.
Source record checked 8 Nov 2026
Meta
Muse app update adding live voice calls and animated, lip-synced avatars, rolling out in the US from November 5, 2026.
Source record checked 8 Nov 2026
Pangram
Pangram's AI-image detector with heatmaps plus a faster Flash model for AI-generated video, via API and browser extension.
Source record checked 8 Nov 2026
Anthropic
Claude beta that links US bank and card accounts via Plaid, read-only, to answer spending and balance questions.
Source record checked 8 Nov 2026
Preferred Networks
Preferred Networks' open-weight translation models in 0.8B-31B sizes covering 53 languages, strongest on Japanese-English.
This entry is marked for review.
Source record checked 8 Nov 2026
Redis
Redis's real-time context engine for AI agents, expanded with intent routing, memory, playbooks and agent messaging.
Source record checked 8 Nov 2026
Amazon
SageMaker real-time endpoints can split one NVIDIA GPU into up to seven isolated MIG partitions shared by several models.
Source record checked 8 Nov 2026
TestGrid
TestGrid's AI operating system for QA: agentic test creation and self-healing on real devices, cloud, on-prem or air-gapped.
Source record checked 8 Nov 2026
athenahealth
athenahealth's MCP server letting US patients connect athenaOne health records to Claude with portal sign-in and revocable access.
Source record checked 7 Nov 2026
BotGauge
Platform that red-teams, traces, evaluates and monitors AI agents with adaptive multi-turn attacks and 200+ attack strategies.
This entry is marked for review.
Source record checked 7 Nov 2026
DigitalOcean
DigitalOcean's native iPhone app for monitoring and approving actions of agents running on DigitalOcean Managed Agents.
Source record checked 7 Nov 2026
Anthropic
Anthropic's open-source zkSNARK prototype proving a hosted model's output came from committed weights without revealing them.
This entry is marked for review.
Source record checked 7 Nov 2026
Google's Gemini API tool for agents that operate user interfaces, now generally available in the Interactions and GenerateContent APIs.
This entry is marked for review.
Source record checked 7 Nov 2026
OpenAI
OpenAI's security-research model, rated Critical for cyber risk, available through API and Codex to select Daybreak Red customers.
Source record checked 7 Nov 2026
AWS
Persistent, named agents inside AWS Kiro Crew that own recurring dev tasks, coordinate worker sessions, and keep inspectable memory.
Source record checked 7 Nov 2026
Microsoft
Microsoft Defender capability, now GA, that finds 50+ local AI agents and their MCP configs on Windows and macOS endpoints.
Source record checked 7 Nov 2026
Anthropic
Anthropic's beta Messages API feature that keeps conversation state server-side so clients send only new messages.
Source record checked 7 Nov 2026
Protegrity
Protegrity's 278M-parameter open-weights encoder that tags 69 PII types in free text, published on Hugging Face.
Source record checked 7 Nov 2026
Ricursive Intelligence
Ricursive's Apache 2.0, 65B dense open-weight model with 1M context, claiming 80.5 on Terminal-Bench 2.1.
Source record checked 7 Nov 2026
Sysdig
Sysdig's runtime control for AI agents that checks each action against policy and can allow, hold for approval, or block it.
Source record checked 7 Nov 2026
Nace AI
Nace AI's open-weight 27B dense decision model that returns probabilities over up to 255 options instead of free text.
This entry is marked for review.
Source record checked 6 Nov 2026
Baidu
Baidu's open-weight 280B MoE base model (14B active), pretrained on 33.5T tokens, with 1M-token context at inference.
Source record checked 6 Nov 2026
Tencent
Tencent's Hy4 LLM family: 770B-parameter (49B active) MoE with 1M context; full release adds visual understanding, with Hy4 Flash for cost.
This entry is marked for review.
Source record checked 6 Nov 2026
OpenAI
OpenAI's Images 2.6 generates images 50% faster; API models Sunburst (quality) and Flare (layout) add SVG export and a complexity parameter.
Source record checked 6 Nov 2026
Thinking Machines Lab
Thinking Machines Lab's 594B-parameter (17B active) MoE model with image input, on Tinker now and Apache 2.0 weights promised.
Source record checked 6 Nov 2026
Meituan
Meituan's LongCat-2.5 combines image understanding with coding for visual coding, design and multi-step codebase tasks; API only so far.
This entry is marked for review.
Source record checked 6 Nov 2026
Pruna AI
Pruna AI's 33B text-to-image and image-editing model derived from MiniMax-H3, with a ComfyUI variant.
This entry is marked for review.
Source record checked 6 Nov 2026
Perplexity
Perplexity's persistent agents that own ongoing work like campaigns or bug triage, with their own desktop, browser and thread; Max only.
Source record checked 6 Nov 2026
Ai2
Ai2's benchmark of how language models track changing entity states across long documents, in four difficulty levels.
Source record checked 6 Nov 2026
Cognition
Cognition's 276B-parameter (12B active) coding model for Devin Desktop and CLI, about 3x faster than SWE-2 at roughly one-fifth the price.
Source record checked 6 Nov 2026
Hugging Face
Hugging Face Swift packages: swappable speech engines behind Apple's SpeechAnalyzer, and typed probability answers from local models.
Source record checked 3 Nov 2026
vLLM Project
Open-weight 26B multimodal decision model from vLLM Semantic Router that returns answer probabilities instead of generated text.
Source record checked 3 Nov 2026
Ai2
Ai2 benchmark, data and code for testing whether LLM judges can tell novel research ideas from less novel ones.
Source record checked 3 Nov 2026
Hugging Face
Coalition initiative, with Hugging Face as infrastructure partner, to release publicly funded science data in AI-ready formats.
Source record checked 3 Nov 2026
Hugging Face
Proposed Hugging Face standard that packages shrunk-model integration tests to run cheaply on CPU in Transformers CI.
Source record checked 3 Nov 2026
Vercel
Vercel Labs' Apache-2.0 Rust CLI giving coding agents desktop app control through OS accessibility trees instead of screenshots.
Source record checked 2 Nov 2026
Akamai
Open-source Spin component that runs agent-generated code in WebAssembly isolation with limits on compute, memory and network.
Source record checked 2 Nov 2026
Akamai
Akamai's managed inference service with OpenAI-compatible APIs, 12 launch models from 8B to 2.8T parameters, and Blackwell GPUs.
Source record checked 2 Nov 2026
Canva
Canva Research's design-editing model that restyles and edits whole designs while keeping every element editable.
Source record checked 2 Nov 2026
DiffuSpace
DiffuSpace's Apache-2.0 diffusion language model: 35B total, about 3B active parameters, block-parallel decoding on a Qwen3.6 backbone.
This entry is marked for review.
Source record checked 2 Nov 2026
Gemini in Google Slides can now edit several slides or a whole deck from one prompt, keeping edits native and undoable.
Source record checked 2 Nov 2026
Google DeepMind
Google DeepMind's real-time interactive world model with native audio, prompt steering and two-minute consistent scenes.
Source record checked 2 Nov 2026
GitHub
GitHub Copilot code review adds a multi-agent Max level in public preview and GA approvals that count toward required reviews.
Source record checked 2 Nov 2026
GitHub
GitHub push protection now uses a context-aware classifier, built with Microsoft, to block secrets with no predefined pattern.
Source record checked 2 Nov 2026
IBM
IBM's generally available control plane that registers AI agents as identities with delegation, task-scoped tokens and audit trails.
Source record checked 2 Nov 2026
GitHub
GitHub Copilot can generate and edit images in VS Code, the Copilot app and CLI chat via GPT Image 2.5 Flare, in public preview.
Source record checked 2 Nov 2026
Instinct (Spear Street Technology)
Invite-only personal AI agent you text or call, with Stripe Link payments, 1Password logins and OpenTable bookings.
Source record checked 2 Nov 2026
InclusionAI
inclusionAI's MIT-licensed 560B hybrid-linear MoE with 25B active parameters and a 1M-token context window.
Source record checked 2 Nov 2026
Microsoft AI
Microsoft AI's image-editing preview model: restyling, relighting, text edits and asset restoration, free in MAI Playground.
Source record checked 2 Nov 2026
The Biological Computing Co. (TBC)
The Biological Computing Co.'s image-to-video model built on Wan 2.2, claimed 5x faster, on fal and OpenRouter.
This entry is marked for review.
Source record checked 2 Nov 2026
Particle
Particle's coding model post-trained from DeepSeek V4.1 Flash to cut reasoning and tool-call tokens in agentic software work.
Source record checked 2 Nov 2026
Phonic
Phonic's speech-to-speech model for agent tasks: 50+ languages, tool calls while the caller speaks, and a claimed #1 on an AA index.
Source record checked 2 Nov 2026
Rime
Rime's speech-to-speech model for enterprise voice agents that writes inspectable text before speaking and can be fine-tuned.
Source record checked 2 Nov 2026
Alibaba (DAMO Academy)
Alibaba DAMO's 16B embodied model pairing an 8B reasoner with an 8B video generator for visual planning.
This entry is marked for review.
Source record checked 2 Nov 2026
GitHub
A Copilot app plugin in public preview for triaging code scanning alerts, getting local AI fixes and running /security-review.
Source record checked 2 Nov 2026
DeepSeek
DeepSeek's MIT-licensed prototype agent that watches a GitHub repo and Feishu chats and answers requests with pull requests.
Source record checked 2 Nov 2026
Anthropic
Anthropic's beta API reporting token usage and cost for Claude Console and Enterprise orgs, grouped by model, workspace, key or user.
Source record checked 2 Nov 2026
1Password
1Password feature that lets AI agents sign in to sites after user approval, without seeing the password or one-time code.
Source record checked 1 Nov 2026
Opera
Opt-in early preview that lets websites and extensions in Opera browsers run AI models locally, with no API key.
Source record checked 1 Nov 2026
Pika
Pika's post-trained MiniMax H3 video model, distilled to 6 steps, with stereo audio and up to 12 reference inputs.
Source record checked 1 Nov 2026
Cohere
Cohere's reranker family in Pro and Fast tiers, with optional passage-level Context Focus and calibrated relevance scores.
Source record checked 1 Nov 2026
Sema4.ai
Sema4.ai 3.0 adds team workspaces, Verified Queries and Logic, and an Agent Inbox to cut enterprise agent token spend.
This entry is marked for review.
Source record checked 1 Nov 2026
TanStack
Framework-agnostic TypeScript AI SDK; 1.0 stabilizes core APIs with 23 provider adapters, while the agent harness stays experimental.
Source record checked 1 Nov 2026
Oracle
Oracle and Google Cloud add a GA remote MCP server, Gemini connector (preview) and bi-directional Iceberg lakehouse for Oracle AI Database.
Source record checked 30 Oct 2026
Oracle
Oracle's goal-driven Fusion Companion and in-app Agentic Home workspace for delegating work to Fusion agents.
Source record checked 30 Oct 2026
Oracle
Oracle's public Git repository making Fusion Applications setup machine-readable so AI agents can plan, apply and test configuration.
Source record checked 30 Oct 2026
Anthropic
Anthropic's program giving researchers free Claude credits, expanded on Oct 28, 2026 to funders, institutes and research projects.
Source record checked 29 Oct 2026
Artificial Analysis
Artificial Analysis's nine-eval model benchmark composite, now 60% private test data, plus a new Progress Index v1.
Source record checked 29 Oct 2026
Anthropic
Anthropic's first results from six drug-developer partnerships, including 35,000 antibody designs and an 89% drop in document-review time.
Source record checked 29 Oct 2026
NVIDIA
CUDA 13.4 memory limits plus an open-source GPU Fractioning Operator let Kubernetes pods share one GPU with enforced memory caps.
Source record checked 29 Oct 2026
Cursor
Cursor's TypeScript SDK for defining agents as version-controlled projects with evals, event triggers and cloud or self-hosted deployment.
Source record checked 29 Oct 2026
Surge AI
Surge AI's benchmark of 50 end-to-end professional tax assignments; the best published model passes 10.8% at a 95% rubric threshold.
Source record checked 29 Oct 2026
Google's cyber- and long-horizon-focused Gemini 4 model, ranked #1 on Artificial Analysis Index v5 (59) with restricted early access.
Source record checked 29 Oct 2026
GitHub
A Copilot feature that watches a pull request and fixes failing Actions workflows and merge conflicts in a cloud environment.
Source record checked 29 Oct 2026
GitHub
An agentic whole-repository vulnerability scanner in GitHub code scanning that covers languages CodeQL does not support.
Source record checked 29 Oct 2026
GitHub
A GitHub code scanning detector for prompt injection, sensitive data disclosure and excessive agency in AI apps, run on pull requests.
Source record checked 29 Oct 2026
GitHub
GitHub's agent-first app that runs Copilot work from issue to merge, with multi-repo Projects, computer use and in-chat image generation.
Source record checked 29 Oct 2026
GitHub
Cloud-run Copilot workflows combining prompts, triggers and allowed tools, shareable across a team and able to learn from outcomes.
Source record checked 29 Oct 2026
GitHub
GitHub's context layer for Copilot agents: cross-client memory, 'dreaming' that proposes skill updates, and cross-repo code search.
Source record checked 29 Oct 2026
GitHub
An install-time policy gate between developer or CI endpoints and npm registries, blocking malicious or risky packages.
Source record checked 29 Oct 2026
Google's preview platform for building one app that runs in both Search and the Gemini app, connected via MCP or Android AppFunctions.
Source record checked 29 Oct 2026
Graphify Labs
Open-source skill that turns a codebase, docs and PDFs into a queryable knowledge graph for 20+ coding agents, with no LLM calls for code.
Source record checked 29 Oct 2026
Paul Bakaus
Apache-2.0 design toolkit that gives coding agents 24 design commands and a 62-rule detector to catch generic AI-generated UI.
Source record checked 29 Oct 2026
GitHub
The programmable interface to the Copilot agent harness, being renamed from GitHub Copilot SDK to Microsoft Copilot SDK.
This entry is marked for review.
Source record checked 2 Nov 2026
NVIDIA
NVIDIA's open ASR model for 40 languages with 5.33% average English WER, released with the 463,062-hour Granary v2 dataset.
Source record checked 29 Oct 2026
NVIDIA
NVIDIA's AI cluster operations software, now with autonomous job and hardware recovery for Kubernetes workloads managed by Run:ai.
Source record checked 29 Oct 2026
GitHub
GitHub Design's MIT-licensed AI prototyping tool that builds interactive prototypes from your own design-system components via Copilot.
Source record checked 29 Oct 2026
NVIDIA
A TensorRT 11.1 feature that builds candidate engines over multiple build routes and picks the fastest one meeting an accuracy criterion.
Source record checked 29 Oct 2026
Microsoft
Microsoft's sandbox and 507-task benchmark that grades agents by final database state, now reusable as RL rewards.
Source record checked 29 Oct 2026
Baidu
Baidu's MIT-licensed 3B OCR model that parses multi-page and book-length documents with near-constant memory via SR-SWA attention.
Source record checked 29 Oct 2026
DeepSeek
DeepSeek's open-source production inference engine for V4.1-Flash, running on NVIDIA Blackwell GPUs and Huawei Ascend 950 NPUs.
Source record checked 28 Oct 2026
Block
Block's full-duplex speech-to-speech model trained from scratch on human conversation, small enough to run on a high-end laptop.
Source record checked 28 Oct 2026
ElevenLabs
ElevenLabs' speech-to-speech models (v1 Mini, v1 Max) with ~570-620 ms to first audio, plus a 90+ language live translation API.
This entry is marked for review.
Source record checked 28 Oct 2026
Gemini in Android Auto can now answer using your Gmail, Calendar, Photos, Maps, Search and YouTube context while you drive.
Source record checked 28 Oct 2026
Google Cloud's A5X instances on NVIDIA Vera Rubin NVL72, announced at GTC Berlin with general availability due in November 2026.
Source record checked 28 Oct 2026
Stanford
Stanford and Radical Numerics' text-to-image models trained from scratch with no language pretraining, using up to 275x less compute.
Source record checked 28 Oct 2026
Liquid AI
Liquid AI's ~2B-class open-weight family: a 1.7B text model with 256K context that fits in about 1 GB at 4-bit.
Source record checked 28 Oct 2026
InclusionAI
InclusionAI's MIT-licensed 124B (5.1B active) hybrid-linear MoE tuned for medical reasoning and healthcare deep research.
Source record checked 28 Oct 2026
Microsoft
Microsoft's orchestration layer that runs LLM agents as a mock physician panel, scoring 83.2% on 304 NEJM cases in a Science paper.
This entry is marked for review.
Source record checked 28 Oct 2026
MemTensor
Memory-native LLMs (36B and 288B MoE) that keep history in learned memory, claiming 66x less storage than a KV cache at 128K tokens.
Source record checked 28 Oct 2026
OpenAI
OpenAI's frontier models and Codex coming to Google Cloud's Model Garden and Marketplace, billable against Google spend commitments.
Source record checked 28 Oct 2026
InternScience
InternScience's ~35B MoE agent model (3B active) scoring 81.7 on BrowseComp and 61.2 on SWE-Bench Pro, with FP8 and GGUF builds.
This entry is marked for review.
Source record checked 28 Oct 2026
Apple
iOS 27.2 gives Siri AI memory of past conversations, new data sources, an app-exclusion setting and five more languages.
Source record checked 28 Oct 2026
Walmart
Walmart's MCP server that lets approved third-party AI agents search inventory and buy for shoppers, with per-agent spending caps.
Source record checked 28 Oct 2026
Agentic AI Foundation
Anthropic donated the Agent Skills open standard to the Linux Foundation's Agentic AI Foundation, with 70+ platforms adopting it.
Source record checked 27 Oct 2026
Hugging Face
Hugging Face's Swift library, a drop-in replacement for Apple's Foundation Models framework across local and hosted models.
Source record checked 27 Oct 2026
Amazon
AWS service that validates code vulnerabilities in a sandbox, ranks them by business impact and opens fix pull requests.
Source record checked 27 Oct 2026
Cloudflare
Cloudflare's first System One extraction model returns typed values, estimates and forecasts in one pass with no output tokens.
Source record checked 27 Oct 2026
Cloudflare
Cloudflare's edge script exposes storefront tools to browser agents via WebMCP; Commerce.com enabled it on nearly 100,000 domains.
Source record checked 27 Oct 2026
AMD
AMD's FlyDSL kernel language is now an optional PyTorch backend on MI350-series GPUs, with up to 3.01x end-to-end vLLM gains.
Source record checked 27 Oct 2026
GitHub
Copilot's LLM-based permission mode auto-approves clearly safe commands, file paths and URLs and asks you about the rest.
Source record checked 27 Oct 2026
GitHub
The GitHub app for Teams now lets you @mention GitHub to have Copilot investigate a bug and open a pull request.
Source record checked 27 Oct 2026
IBM
IBM's Apache-2.0 MLIR scheduler turns KTIR kernels into dataflow pipelines and SpyreCode for the Spyre accelerator.
Source record checked 27 Oct 2026
NVIDIA
Open BioNeMo protein-design model built with the Baker Lab; NVIDIA claims ~13x faster generation than RFdiffusion3.
Source record checked 27 Oct 2026
NVIDIA
NVIDIA Confidential Computing option that lets frontier model builders run sensitive workloads on third-party bare-metal GPUs.
Source record checked 27 Oct 2026
NVIDIA
NVIDIA's Cosmos Data Factory blueprints and Cosmos 3 tools for generating, curating and evaluating physical-AI training data.
This entry is marked for review.
Source record checked 27 Oct 2026
NVIDIA
NVIDIA's CUDA-X sparse direct solver library, now generally available, with 4-10x reported speedups in engineering simulation.
Source record checked 27 Oct 2026
NVIDIA
NVIDIA's regional StormScope EU nowcasting model, trained on European data, joins open Earth-2 weather and climate models.
Source record checked 27 Oct 2026
NVIDIA
NVIDIA blueprint for building always-on factory manager agents and specialist inspection and root-cause agents; early access on GitHub.
Source record checked 27 Oct 2026
NVIDIA
NVIDIA's AIOps platform that correlates GPU, network and job telemetry in a knowledge graph to find and fix AI factory faults.
Source record checked 27 Oct 2026
Revolut
Revolut's in-production transaction foundation model, serving 80 million customers on NVIDIA Blackwell Ultra GPUs.
Source record checked 27 Oct 2026
Vercel
Vercel Agent now investigates monitor alerts and forwards only Critical ones to paging tools such as incident.io.
Source record checked 27 Oct 2026
NVIDIA
NVIDIA's GTC Berlin reference humanoid on Isaac GR00T with 31 degrees of freedom, due early 2027.
Source record checked 26 Oct 2026
General Reasoning
Research model series that trains a CEO-plus-specialists agent organisation as one unit, built on Nemotron 3 Super.
Source record checked 26 Oct 2026
Anthropic
Anthropic-run Claude API access inside Google Cloud with IAM access control, Marketplace billing and Opus 5.5 fast mode.
Source record checked 26 Oct 2026
IQuest
320B sparse MoE (15B active) agentic-CLI model with 512K context and 83.2 on Terminal-Bench 2.1.
This entry is marked for review.
Source record checked 26 Oct 2026
Kling AI
Kuaishou's Kling 4.0 video model generates up to 30 seconds in one pass with 15-asset Omni Reference.
Source record checked 26 Oct 2026
Nanbeige
33B dense open-weight reasoning and coding model with a 512K context window, released under Apache 2.0.
Source record checked 26 Oct 2026
Microsoft
Microsoft's agentic-insider defense: discovers 60 agent types, scores risk and uses Guardian Agents to supervise other agents.
Source record checked 26 Oct 2026
StepFun
StepFun's agent-first smartphone running Step AOS 1 and the Amoo agent, on sale in China from Oct 16 at RMB 3,999.
Source record checked 26 Oct 2026
VideoGen
Text- and image-to-video model producing multi-shot videos up to 10 minutes with joint audio-video generation.
Source record checked 26 Oct 2026
OpenAI
ChatGPT's deck builder in Space: create, edit, co-author, present, and export slides to PowerPoint, Google Slides, or PDF.
Source record checked 23 Oct 2026
Anthropic
Anthropic's proactive agent app on Claude Opus 5.5 that sends a morning brief, monitors connected apps, and asks approval before acting.
Source record checked 23 Oct 2026
NVIDIA
NVIDIA's 4B open looped model whose compute is set at serving time via loop count (1, 2, or 4), with 256K context.
This entry is marked for review.
Source record checked 23 Oct 2026
Rockwell Automation
Rockwell's AI-native orchestration platform that links agents, skills, and workflows across the industrial automation lifecycle.
Source record checked 23 Oct 2026
Recursive
Recursive's Apache 2.0 40B mixture-of-experts coding model with 2.5B active parameters, built to run locally.
Source record checked 23 Oct 2026
OpenAI
OpenAI's GPT-6 Pro in ChatGPT now splits a task across several agents that run at once, with faster responses and better tool use.
This entry is marked for review.
Source record checked 23 Oct 2026
IBM
IBM's Apache 2.0 speech models: a 2B ASR/translation model at 5.06% English WER and a 470M Turbo CTC model at 19,870 RTFx.
Source record checked 23 Oct 2026
Intel
Intel's Xe3P inference accelerator (formerly Crescent Island): up to 480GB LPDDR5X, 1.3 PFLOPS FP8, 350W air-cooled PCIe.
This entry is marked for review.
Source record checked 23 Oct 2026
Kognitos
61 role-based finance and banking agents backed by 515 SOPs, plus 133 ready-made Context Graph monitors and an SOP-drafting agent.
Source record checked 23 Oct 2026
Baseten
Baseten's open decision models (27B and 9B) that return probabilities for choice, score, and yes/no questions instead of text.
This entry is marked for review.
Source record checked 23 Oct 2026
Ai2
Ai2's open Earth-observation model with precomputed 10 m global embeddings and a no-login Embeddings Explorer.
Source record checked 23 Oct 2026
Oracle
Oracle's no-code and low-code studio for building, testing, and managing industry-specific agents, with MCP tool access.
This entry is marked for review.
Source record checked 23 Oct 2026
Amazon
AWS's Apache 2.0 native PyTorch backend for Trainium2 and Trainium3: .to('neuron'), eager, torch.compile, and NKI kernels.
Source record checked 23 Oct 2026
PyTorch
PyTorch-native pre- and post-training platform; the October 2026 update adds TitanRL, Kimi K3, DeepSeek-V4 and Muse Spark Open recipes.
Source record checked 23 Oct 2026
Open-source Google and Meta PyTorch backend for TPUs: eager and torch.compile modes, ~450 ATen ops, model.to('tpu').
Source record checked 23 Oct 2026
Mistral AI
Mistral's speech family: transcription from $0.0035/min, a multilingual TTS model, and a LiveKit-based voice agent SDK.
This entry is marked for review.
Source record checked 23 Oct 2026
ThreatLocker
ThreatLocker's external policy layer that limits what AI agents can access, execute, and connect to, built on Allowlisting.
Source record checked 23 Oct 2026
DeepSeek
DeepSeek's web Search and Fetch endpoints for agents, billed at $0.75 per 1,000 successful calls each.
Source record checked 22 Oct 2026
Motive
Motive's fleet agents act on telematics alerts by calling drivers, opening work orders and escalating to managers.
Source record checked 22 Oct 2026
Notion
Notion's rebuilt agents run single tasks for 24+ hours, with an Auto model router and a two-pool credit system.
Source record checked 22 Oct 2026
NVIDIA
NVIDIA's models and libraries for grounding agents in enterprise data: Nemotron retrievers, Kumo models and Auto-Ontology.
Source record checked 27 Oct 2026
NVIDIA
NVIDIA's modular Vera Rubin AI factory reference design, cutting on-site deployment from six months to 14 days.
Source record checked 27 Oct 2026
Inception
Inception's flagship diffusion LLM: ~1,068 tokens/sec, 1M-token context, $0.20/$0.70 per million tokens.
Source record checked 20 Oct 2026
MiniMax
MiniMax's low-cost agentic and coding model, priced at $0.15/$0.60 per million tokens, officially rolled out Oct 19, 2026.
Source record checked 20 Oct 2026
Vercel
Vercel's public-beta runtime for continuous background workers (Celery, Dramatiq, Temporal), announced Oct 15, 2026.
Source record checked 20 Oct 2026
Vercel
Persistent storage for Vercel Sandbox that keeps agent code, data and memory across runs; GA as of Oct 15, 2026.
Source record checked 20 Oct 2026
ServiceNow
ServiceNow's signed runtime credential that defines and enforces what each AI agent or MCP server may do, with a kill switch.
Source record checked 19 Oct 2026
LightOn
LightOn's open ~2B local code-search agent that cut Claude Code API tokens by 34% versus grep in its own tests.
Source record checked 19 Oct 2026
IBM
IBM's 8B and 30B Granite models bundling decision, safety guardian and speech-to-text adapters in one endpoint with shared KV cache.
Source record checked 3 Nov 2026
Ai2
Ai2's climate emulator pairing 100 km ACE with a HiROv2 downscaler to produce global 3 km data at ~1/300th the cost of X-SHiELD.
Source record checked 19 Oct 2026
FrontiersMind
FrontiersMind's 27B Apache-2.0 decision model that scores supplied options in one forward pass, reporting 90.7% vs TypeSafe Jev's 83.9%.
Source record checked 19 Oct 2026
Mistral AI
Mistral's first time-series forecasting models: four CPU-friendly sizes (5M-209M) with probabilistic output and open weights.
Source record checked 19 Oct 2026
DMIS Lab
Apache-2.0 English encoder family (150M, 380M) with 32K native context that reports beating ModernBERT, plus a 150M embedder.
Source record checked 19 Oct 2026
Fish Audio
Fish Audio's open-weight 35B MoE text-to-speech model with bracketed stage directions across 83 languages.
Source record checked 18 Oct 2026
Black Forest Labs
Black Forest Labs' 5.6B open-weights image model for text-to-image, editing, and up to 10 reference images, running on 8 GB GPUs.
Source record checked 18 Oct 2026
Elorian
Elorian's first model: a visual reasoning model that crops, pulls video frames, and runs code on the source image while it thinks.
Source record checked 18 Oct 2026
Google's experimental on-device macOS meeting note-taker, running EmbeddingGemma 2 and Gemma 4 locally on Apple Silicon.
Source record checked 18 Oct 2026
Apple
Apple's smart home hub with a 7-inch square display and A18 chip, built around Siri AI; ships November 13, 2026.
This entry is marked for review.
Source record checked 18 Oct 2026
Tencent
Tencent's open-source agent memory stack: a 30B-A3B memory-writer model, an SDK, and the MemHub local memory app.
Source record checked 18 Oct 2026
Krea
Krea's image-editing model: 2K edits in under 8 seconds, region and bounding-box control, up to six references, and LoRA fine-tuning.
Source record checked 18 Oct 2026
Liquid AI
Liquid AI's largest LFM2.5 model: a 24B MoE with 2B active parameters, 131K context, built for on-device agents.
Source record checked 18 Oct 2026
Microsoft
Microsoft AI's 2.8T-parameter MoE base model, pre-trained on 16K GB200 GPUs in 38 days; in post-training, not yet released.
Source record checked 18 Oct 2026
ScaleDown
ScaleDown's small non-generative models for classification, extraction, and summarization that return only allowed outputs.
This entry is marked for review.
Source record checked 18 Oct 2026
Paper Instruments
Paper Instruments' GLM-5.3 Flash post-trained model for finance, law, consulting, and operations deliverables.
Source record checked 18 Oct 2026
Vivix AI
Vivix AI's streaming-native video model for worlds that change in real time from text, voice, and motion input.
This entry is marked for review.
Source record checked 18 Oct 2026
Zenzap
AI-native workplace messenger combining group chat, tasks, and calendars with AI agents reachable by text or voice.
Source record checked 18 Oct 2026
NVIDIA
NVIDIA's open 120B mixture-of-experts multimodal agent model with 12B active parameters, released with weights, data and recipes.
Source record checked 17 Oct 2026
Perplexity
Perplexity's open-source Firecracker microVM sandbox platform that runs every Perplexity Computer agent task in isolation.
Source record checked 17 Oct 2026
Sakana AI
Sakana AI's Apache 2.0 pipeline that orchestrates coding agents to turn a task.md file into experiments, ablations and a paper.
Source record checked 17 Oct 2026
Microsoft
Microsoft says its Maia 200 AI accelerator is in production serving Copilot, and unveiled Maia 300 for 2027 deployment.
Source record checked 16 Oct 2026
Cisco
Cisco's open-source DefenseClaw gains an Edge Connector that allows or blocks AI agent tool calls directly on IoT and robot hardware.
Source record checked 16 Oct 2026
Meta
Meta's open-weight 816B-parameter MoE model (42B active, 1M context) for coding and long-horizon agent work, released Oct 15, 2026.
Source record checked 16 Oct 2026
NVIDIA
NVIDIA's open 74B MoE model (9B active) packed into 24GB for local agents on RTX Spark; released Oct 15, 2026.
Source record checked 16 Oct 2026
NVIDIA
NVIDIA's free, open-source personal AI router that spreads local model workloads across PCs on your network; out of beta Oct 13, 2026.
Source record checked 16 Oct 2026
Agent37
YC F26 startup selling low-cost agent sandboxes (gVisor or full VM) with prebuilt harnesses for Claude Code, Hermes and opencode.
Source record checked 15 Oct 2026
Amazon
Amazon's generally available real-time speech-to-speech model in Bedrock, with a 256K context window and asynchronous tool calling.
Source record checked 15 Oct 2026
Cisco
Cisco's agentic customer-experience harness for the Webex AI Agent platform, keeping agents working after the interaction ends.
Source record checked 15 Oct 2026
DigitalOcean
DigitalOcean's managed data layer for agents: ingestion flows, a catalog with vectors, and SQL or vector query, in private preview.
Source record checked 15 Oct 2026
DigitalOcean
DigitalOcean's feedback loop for agents: Insights logs activity, Signals flags failures, Evaluations & Simulations tests fixes. Public preview.
Source record checked 15 Oct 2026
Z.ai
Z.ai's legal agent model built on GLM-5.3-Flash, scoring 19.2% all-pass on Harvey's LAB benchmark.
This entry is marked for review.
Source record checked 15 Oct 2026
xAI
xAI's speech-to-speech model in the Voice Agent API, priced at $0.08/min and scoring 86.9 on Artificial Analysis' Speech-to-Speech Index.
Source record checked 15 Oct 2026
Noiz AI
Real-time interactive world model from Noiz AI and HKUST that generates explorable video with spatial stereo audio from one image.
Source record checked 15 Oct 2026
Ironclad
Ironclad's conversational contract agent, grounded by a Contract Knowledge Graph of clauses, obligations and past decisions.
Source record checked 15 Oct 2026
Institute of Foundation Models
Six open Arabic-focused language models from 900M to 375B parameters with 512K context, plus Jais Speech for Arabic recognition.
Source record checked 15 Oct 2026
Sqwish Labs
Sqwish Labs' decision-model family returning answer probabilities in one pass, with open 3.7B and 900M models and a fine-tuning API (OneHot).
Source record checked 15 Oct 2026
Microsoft
Microsoft's Qwen3.5-9B-based decision-scoring model returns calibrated option probabilities in one pass, at $0.042 per million input tokens.
Source record checked 15 Oct 2026
Meta
Meta's most capable Muse models, with active memory and long-goal focus for agents, 1M context and $4/$20 per million tokens.
Source record checked 15 Oct 2026
Perplexity
Perplexity's late-interaction multimodal embedding models (0.6B and 9B) with 128-dim per-token vectors, open on Hugging Face.
Source record checked 15 Oct 2026
Reka
Reka's 19B omni-reasoning model that handles text, images, video and robot actions in one network, in research preview.
Source record checked 15 Oct 2026
Workday
Workday's generally available environment for building enterprise agents with isolated runtimes, memory and governed HR/finance data access.
Source record checked 15 Oct 2026
Databricks
Databricks' no-code system for building and auto-optimizing domain-specific AI agents, now generally available with synthetic eval-data generation and multi-agent orchestration.
Source record checked 14 Oct 2026
Amazon
Amazon's third-generation Nova frontier model family, unveiled at an AWS keynote with a 1M-token context window and a dedicated agentic tool-use variant.
Source record checked 14 Oct 2026
OpenAI
General availability of autonomous Agent Mode in OpenAI's ChatGPT Atlas browser, letting ChatGPT complete multi-step web tasks like bookings and checkout on a user's behalf.
Source record checked 14 Oct 2026
Anthropic
Anthropic's public marketplace for Claude Agent Skills, letting developers publish, monetize, and install reviewed Skills directly into Claude Code and the Claude apps.
Source record checked 2 Nov 2026
Cursor
Cursor's second-generation in-house coding model, now the default agent model in the editor, with faster multi-file edits and native terminal tool use.
Source record checked 14 Oct 2026
Cognition
Cognition's autonomous software engineering agent, now able to run unsupervised multi-day engineering sprints with persistent repo memory and automatic PR-review handling.
Source record checked 14 Oct 2026
Lakera
Lakera's real-time guardrail API for LLM apps, updated with multi-agent protection and a new benchmark for detecting indirect prompt injection in tool outputs.
Source record checked 14 Oct 2026
Meta
Meta's open-source SDK for building and orchestrating multi-agent systems on top of its Muse model family.
Source record checked 14 Oct 2026
NVIDIA
NVIDIA's next-generation rack-scale AI system pairing Vera CPUs with Rubin GPUs, detailed with a shipping timeline pulled forward to compete with AMD's newly announced Helion rack.
This entry is marked for review.
Source record checked 22 Oct 2026
AMD
AMD's open rack-scale blueprint for agentic AI infrastructure, with CPU-only configurations up to 32,768 EPYC cores and an integrated model-call token router.
Source record checked 13 Oct 2026
Arena
Community-voting LLM leaderboard (formerly LMArena/Chatbot Arena) that now sells agent-evaluation analytics and a real-world alignment benchmark to AI labs.
Source record checked 13 Oct 2026
Open-weight 740M-parameter multimodal embedding model built on Gemma 4 that encodes text, images, audio, and video for on-device search.
Source record checked 13 Oct 2026
Google's updated text-to-image and photo-editing model with better text rendering, character consistency, and roughly half the per-image price of its predecessor.
Source record checked 13 Oct 2026
Manus
General-purpose AI agent that automates multi-step browser and workflow tasks, now operating independently after a $500M+ round at a $4B valuation.
Source record checked 13 Oct 2026
REA
MIT-licensed MCP/CLI toolkit that lets coding agents reverse-engineer apps and native binaries locally, which hit #1 on GitHub Trending in October 2026.
Source record checked 13 Oct 2026
Marvell
Marvell's second-generation CXL 3.2 memory expanders, the industry's first six-channel design, supporting up to 9TB of pooled memory per controller for AI inference.
Source record checked 13 Oct 2026
Cloudflare
Cloudflare's mixture-of-experts decision model that reads text, JSON, images, audio, and video in one request and returns a probability for every allowed answer.
Source record checked 12 Oct 2026
O'Reilly
O'Reilly's enterprise toolkit that grounds AI coding and work assistants in cited, dated practitioner knowledge via an MCP-compatible connector.
Source record checked 12 Oct 2026
Sierra
Sierra's model that listens to a live phone call and scores whether the caller is an AI agent rather than a human.
Source record checked 12 Oct 2026
xAI
xAI's faster, cheaper tier of Grok Imagine Video 1.5, generating 1-15 second clips with native audio from text or image prompts.
Source record checked 12 Oct 2026
JetBrains
JetBrains' open 12B mixture-of-experts coding model, retrained with reinforcement learning to explore, edit, and verify its own code changes for fast local coding agents.
Source record checked 12 Oct 2026
Liquid AI
Liquid AI's open-weight decision models that return a calibrated yes/no/choice/score answer in a single forward pass instead of generating tokens.
Source record checked 18 Oct 2026
Sierra
An open protocol led by Sierra and Meta, nicknamed Poppy, defining how personal AI agents acting on a consumer's behalf interact with businesses.
Source record checked 12 Oct 2026
Perplexity
Perplexity's updated open-weight multimodal decision model, half the price of v1 and top-scoring on the Hugging Face Decision Index 0.3 benchmark.
Source record checked 18 Oct 2026
Promptfoo
Open-source LLM testing, red-teaming, and vulnerability-scanning CLI, updated in v0.124.0 with Claude Opus 5.5 support and new filtering options.
Source record checked 12 Oct 2026
Reka AI
Reka AI's 7B open-weight vision-language model built for on-device agentic use on Jetson, Apple Silicon, and phones.
Source record checked 12 Oct 2026
Anthropic
Anthropic's long-term program giving Claude models, engineers, and threat research to defenders of critical infrastructure and open-source software.
Source record checked 10 Oct 2026
Reflection AI
Reflection AI's first open-weight model, a 501B-parameter MoE positioned as a lower-compute Western answer to Chinese open models.
Source record checked 20 Oct 2026
Anthropic
Anthropic's beta tools for live, Claude-built data dashboards and code-driven animations exported as MP4.
Source record checked 10 Oct 2026
Anthropic
Anthropic's cheapest Claude 5.5-family model, cutting small-prompt pricing up to 90% versus Haiku 4.5 with a 1M-token context.
Source record checked 10 Oct 2026
Google Cloud's enterprise agent for persistent work across connected business systems.
Source record checked 16 Oct 2026
Microsoft
OS-level containment for AI agents on Windows, letting admins define and enforce at runtime which files and networks an agent can access.
Source record checked 10 Oct 2026
Runway
Runway's first open-weight world-action model, turning its video-pretraining expertise into cross-embodiment robot control.
Source record checked 10 Oct 2026
Microsoft
Microsoft's Nvidia RTX Spark Superchip-powered Surface devices, built to run 120B+ parameter AI models locally.
Source record checked 10 Oct 2026
Google's public tool for checking whether an image, video, or audio file carries a SynthID AI watermark, now open globally in English.
Source record checked 10 Oct 2026
BlackFog
BlackFog's endpoint-level AI security layer adding seven layers of prompt-injection protection and agentic AI governance.
Source record checked 7 Oct 2026
OpenAI
OpenAI's new ad format that displays a separate, clearly labeled commercial image alongside ChatGPT's generated image results.
Source record checked 7 Oct 2026
Oracle
Oracle's governed agentic execution runtime, embedded natively in Fusion Applications, launching with 25 new Claw-powered apps.
Source record checked 7 Oct 2026
Synopsys
A frontier AI model co-developed by Synopsys and OpenAI to operate EDA tools directly for chip design and verification.
This entry is marked for review.
Source record checked 7 Oct 2026
Mistral AI
Mistral's 1T-parameter multimodal MoE flagship; open weights published on Hugging Face under a modified-MIT license.
Source record checked 30 Oct 2026
Cohere
Cohere's rebuilt enterprise agent platform adding cross-session agent memory, redesigned orchestration, and per-user token spending caps.
Source record checked 7 Oct 2026
Thales
Thales' add-on that hardens compiled software against AI-assisted reverse engineering and automated vulnerability discovery, with no source changes.
Source record checked 7 Oct 2026
Vega
Vega's agentic cyber-defense platform update with a security-trained model, persistent SOC memory, and SIEM-free log search.
Source record checked 7 Oct 2026
OpenAI
OpenAI's terminal-based coding agent, refreshed at DevDay 2026 with voice control, a multi-task /agents view, reusable cloud environments, and ChatGPT desktop code review.
Source record checked 6 Oct 2026
NVIDIA
NVIDIA's open hardware-and-software stack for containing AI agents, pairing a CPU-level runtime with a DPU watchdog that can quarantine rogue agents in milliseconds.
Source record checked 22 Oct 2026
Alibaba
Alibaba Cloud's enterprise-grade agent lifecycle platform, unveiled at Apsara Conference 2026 as the centerpiece of its 'Agentic Cloud' strategy.
Source record checked 4 Oct 2026
Ema
Ema's platform of autonomous 'AI Employee' agents automating HR, IT, and finance workflows, which just raised a $77M Series B quadrupling its valuation.
Source record checked 4 Oct 2026
Cloudflare
Cloudflare's open-weight 27B decision model that returns typed, calibrated choices instead of text, positioned against TypeSafe's Jev.
Source record checked 4 Oct 2026
Cloudflare
Cloudflare's lightweight 9B sibling to Clef, built for sub-40ms decision-model inference.
Source record checked 4 Oct 2026
Microsoft
Microsoft AI's first real-time streaming speech-to-text model, ranked #1 for transcript accuracy on Artificial Analysis across 60 languages.
Source record checked 4 Oct 2026
Microsoft
Microsoft AI's text-to-speech model covering 23 languages with a consistent voice identity, plus a low-latency Flash variant.
Source record checked 4 Oct 2026
Amazon
AWS Strands Labs' open-weight 2B decision model that picks among predefined options in roughly 100ms on a single consumer GPU.
Source record checked 4 Oct 2026
ServiceNow
ServiceNow CoreAI's framework that turns an enterprise agent's capability gaps into targeted synthetic training data.
Source record checked 3 Oct 2026
OpenAI
A global ChatGPT shopping feature that overlays clothing onto a user's photo and lets shoppers save favorite finds, powered by Images 2.5.
Source record checked 3 Oct 2026
Anthropic
Anthropic's FedRAMP High authorized environment bringing Claude's coding and agentic capabilities to U.S. federal and state agencies.
Source record checked 29 Oct 2026
OpenAI
A GPT-6 Luna-powered endpoint that picks among a developer-defined set of answers in ~150ms instead of generating free text.
Source record checked 23 Oct 2026
Google's new frontier model for long-horizon coding, enterprise knowledge work, and autonomous cyber defense, rolled out first to trusted defenders.
Source record checked 15 Oct 2026
Google's open-source skills pack that lets AI coding agents autonomously find, reproduce, and patch vulnerabilities in a repository.
Source record checked 3 Oct 2026
Inception
Inception's diffusion language model tuned for voice agents, returning a first answer token in under 320ms median while reasoning and calling tools.
Source record checked 3 Oct 2026
Meta
A global Meta Neural Band feature that reads surface electromyography (sEMG) signals to let people handwrite text on any flat surface.
Source record checked 3 Oct 2026
Ai2
Ai2's open, redesigned mixture-of-experts training stack, benchmarked past one trillion total parameters at 2.7x the throughput of its prior FSDP implementation.
Source record checked 3 Oct 2026
Hugging Face
Hugging Face's objective-metric leaderboard for comparing multilingual text-to-speech and voice-cloning models at scale.
Source record checked 3 Oct 2026
OpenAI
A beta marketplace letting eligible enterprise customers spend part of their existing OpenAI commitment on 32 launch partners' software.
Source record checked 3 Oct 2026
Qwen
A community fine-tune of Qwen3.8-27B that fixes effort-ordered reasoning so higher reasoning effort reliably costs more tokens and does better, not worse.
This entry is marked for review.
Source record checked 3 Oct 2026
OpenAI
OpenAI's premium speed tier delivering up to 8x faster token generation (300 tokens/sec) in Codex and up to 6x faster in the API.
Source record checked 3 Oct 2026
Android Studio Rabbit 2 Canary lets developers connect ACP coding agents to IDE build, preview, SDK, and emulator tools.
Source record checked 1 Oct 2026
OpenAI
A collaborative ChatGPT workspace for files and Pages that people and agents can draft, revise, and edit.
Source record checked 1 Oct 2026
OpenAI
A cloud service for scanning connected repositories and reviewing security findings and proposed fixes.
Source record checked 16 Oct 2026
CoreWeave
CoreWeave's unified development layer connecting model/agent inference, observability, data curation, training, and evaluation in one loop.
Source record checked 1 Oct 2026
OpenAI
Always-on ChatGPT agents that pursue ongoing goals across connected apps and report back for review.
Source record checked 1 Oct 2026
Axelera AI
Axelera AI's Europa AI Processing Unit delivers up to 629 TOPS for enterprise and data-center edge inference.
Source record checked 1 Oct 2026
Galiark
Galiark's deterministic AI platform gives U.S. healthcare claims adjudicators auditable, reproducible decisions instead of probabilistic outputs.
Source record checked 1 Oct 2026
Reusable Markdown-based instructions for Gemini and Workspace, with separate rollouts starting October 5 and 13, 2026.
Source record checked 1 Oct 2026
GitHub
GitHub Security Lab's agent workflow for C/C++ fuzzing, coverage analysis and crash triage.
Source record checked 16 Oct 2026
H
H's open-weight Holo4 models combine GUI control, coding, MCP, and API calls across desktop, web, and Android tasks.
Source record checked 1 Oct 2026
Morae
Morae's managed legal-spend review service combining AI invoice analysis with expert oversight.
Source record checked 16 Oct 2026
OpenObserve
OpenObserve's v1.0 release brings agent tracing, LLM monitoring, evaluation, and session annotation into its existing logs/metrics/traces platform.
Source record checked 1 Oct 2026
OpenAI
A developer sign-in integration that can let eligible users apply ChatGPT plan usage inside third-party apps.
Source record checked 1 Oct 2026
VoiceStudio
VoiceStudio is a fully-local, open-source ElevenLabs alternative for voice cloning, dubbing, dictation, and transcription in 646 languages.
Source record checked 1 Oct 2026
Outerlimit
Decentralized zero-trust security layer that authorizes and binds AI agent actions at the moment of tool execution.
Source record checked 30 Sept 2026
Dataiku
A standalone Dataiku product that inventories AI agents across platforms like AWS, Databricks, Google, Microsoft, Salesforce, and Snowflake, and tiers them by risk.
Source record checked 29 Sept 2026
Ando
An agent-native team messaging platform, built as a Slack alternative where AI agents participate as first-class members alongside people.
Source record checked 29 Sept 2026
Anthropic
Anthropic's faster, cheaper Claude 5.5-family model, generating output 30% faster than Sonnet 5 at unchanged pricing.
Source record checked 29 Sept 2026
ElevenLabs
ElevenLabs' new flagship voice model family, ranked No. 1 on Artificial Analysis's Voice TTS Arena, with a low-latency Turbo variant for live calls.
Source record checked 29 Sept 2026
Apple
Apple's M5 Ultra chip, shipping in the Mac Studio from September 22, 2026, with up to 4.3x the AI compute performance of M3 Ultra.
Source record checked 29 Sept 2026
Meta
Meta's new enterprise AI division bundling Business Agent, a Muse API, and the Muse Code coding agent for business and developer customers.
Source record checked 29 Sept 2026
Meta
A keychain-size wearable from Meta with a 2-inch screen, dual cameras, and standalone 5G, letting users talk to Muse without a phone.
Source record checked 29 Sept 2026
Meta
Meta's terminal coding agent, built on Muse Spark 1.2, now bundled into the new Meta Enterprise Platform for business customers.
Source record checked 29 Sept 2026
Google's research moonshot to run TPU AI chips on solar-powered satellites in orbit, with a prototype satellite scheduled to launch October 1, 2026.
This entry is marked for review.
Source record checked 29 Sept 2026
Strada
A no-code browser-automation capability from insurance AI vendor Strada that lets agents work inside carrier portals and legacy systems with no API.
Source record checked 29 Sept 2026
Fastly
Fastly's edge services for filtering AI requests and governing model routing, credentials and spend.
Source record checked 16 Oct 2026
Cloud Range
A contained SOC environment for red-teaming AI security agents against human defenders, paired with a 5-step PROVE readiness framework.
Source record checked 28 Sept 2026
GitHub
GitHub's terminal AI coding agent, updated to v1.0.88 with OSC 777 terminal notifications, safer approval persistence, and MCP reliability fixes.
Source record checked 28 Sept 2026
Apple
Apple's first open vision-language model, built on Qwen3.5-9B, for efficient long-document understanding via selective decompression.
Source record checked 28 Sept 2026
LLM Gateway
A beta routing mode on LLM Gateway that classifies request difficulty and sends prompts to a cheaper or frontier model from a developer-selected pool.
Source record checked 28 Sept 2026
Proofpoint
Proofpoint's unified system reasoning across AI and data risk in one shared graph, using three autonomous agents to detect, investigate, and remediate threats.
Source record checked 25 Sept 2026
AgentOS
An open-source 'Open Agent Operation System' for running and managing agents, now shipping a native macOS app alongside its engine.
Source record checked 25 Sept 2026
Agno
An open-source Python framework for building AI agents and multi-agent teams with knowledge retrieval, MCP support, and multi-provider models.
Source record checked 25 Sept 2026
Bespoke Labs
A Qwen3.5-9B adapter for scoring predefined choices, booleans and rubric levels.
Source record checked 16 Oct 2026
Google's AI-generated summaries and direct answers for natural-language questions typed into Gmail's search bar, now available globally.
Source record checked 25 Sept 2026
Nous Research
Nous Research's open-source agent with persistent memory, reusable skills and messaging-platform access.
Source record checked 16 Oct 2026
Liquid AI
An experimental ~280M-parameter draft model that speeds up LFM2.5-VL-3B vision-language inference up to 3.13x via speculative decoding.
Source record checked 25 Sept 2026
llama.cpp
The v0.5.0 release of the widely used C/C++ LLM inference engine, adding new model support and backend performance work.
This entry is marked for review.
Source record checked 25 Sept 2026
OpenHands
OpenHands' open-source autonomous coding-agent platform, with v1.23.0 adding a universal macOS DMG and new UI themes.
Source record checked 25 Sept 2026
Amazon
Amazon's AI agent for marketplace sellers, now with persistent memory across sessions and always-on autonomous restocking/pricing workflows.
Source record checked 25 Sept 2026
Unsloth
September updates to Unsloth's local model toolkit, covering skills, model support, Docker and multi-user operation.
Source record checked 16 Oct 2026
D-Robotics
D-Robotics AI Lab's open robot simulator that autoregressively generates video observations from joint-trajectory input at 24 FPS.
Source record checked 25 Sept 2026
OpenRouter
OpenRouter's asynchronous API for bundling requests across 70+ models at roughly half the standard per-token price.
Source record checked 24 Sept 2026
Anthropic
Anthropic's first 5.5-family model, matching Claude Fable 5.1 on most work at 40% lower cost.
Source record checked 24 Sept 2026
Base Compute
Base Compute's open-source leaderboard and CLI for benchmarking local inference across models, quantizations, chips, and runtimes.
Source record checked 24 Sept 2026
EvalEval Coalition
A shared reporting schema and open platform, built with the UK AI Security Institute, for making AI benchmark results comparable across setups.
Source record checked 24 Sept 2026
Black Forest Labs
A 7B-parameter open-weight world action model that turns camera frames and text instructions into the next 2 seconds of robot actions.
Source record checked 24 Sept 2026
Google's most expressive text-to-speech model, generating custom voices from natural-language prompts across 100+ languages.
Source record checked 24 Sept 2026
Google's lower-cost sibling to Gemini 3.8 Flash TTS, bringing expressive voice generation to Google Vids.
Source record checked 24 Sept 2026
GitHub
GitHub's CLI extension that compiles Markdown-defined AI automation into standard, sandboxed GitHub Actions workflows.
Source record checked 24 Sept 2026
Intrinsic
Intrinsic's open-source, ROS-compatible building blocks for industrial robot applications.
Source record checked 16 Oct 2026
JetBrains
JetBrains' cross-vendor system for coordinating agentic software development inside and beyond its IDEs.
Source record checked 24 Sept 2026
MiniMax
MiniMax's open-source terminal coding agent, usable with MiniMax models or any OpenAI/Anthropic-compatible provider.
Source record checked 24 Sept 2026
NVIDIA
A 100M-parameter open-weight speaker diarization model handling up to 8 speakers for both live and recorded audio.
Source record checked 24 Sept 2026
Rabbit
Rabbit's cloud agent interface for delegating work across connected computers and devices.
Source record checked 16 Oct 2026
Hugging Face
Hugging Face's open-source repository-memory server that lets coding agents query GitHub issue/PR history alongside a working clone.
Source record checked 24 Sept 2026
StepFun
StepFun's 600B-parameter (27B active) sparse MoE flagship for long-horizon agentic and coding work, priced at $1/M input tokens.
Source record checked 24 Sept 2026
Hugging Face
A release-candidate rewrite of Hugging Face's Tokenizers library that is 3-30x faster while producing identical token IDs to v0.23.
Source record checked 24 Sept 2026
Alibaba
Alibaba's AI training and inference chip with 216GB of memory, built to support 500,000-chip superclusters for future Qwen models.
Source record checked 24 Sept 2026
Google's Kotlin/Android agent framework reaches 1.0 with feature parity to ADK Core plus on-device and Android-first extensions.
Source record checked 23 Sept 2026
Mistral AI
Mistral's multi-step retrieval layer that lets models search, open, navigate, read, and grep documents iteratively instead of one-shot RAG.
Source record checked 23 Sept 2026
Skynet (skynet.io)
Skynet.io's task-executing AI Agents plus a persistent Database that gives them shared, cross-run memory.
This entry is marked for review.
Source record checked 23 Sept 2026
OpenAI
OpenAI's managed API bringing Codex's agent harness, durable sessions, and hosted sandboxes to developers building production agents.
Source record checked 1 Nov 2026
NIST
NIST's sequestered testbed for blind-data evaluation of AI models, launching with vision-language tasks in quantum science, genomics, and public safety.
Source record checked 23 Sept 2026
Salesforce
Salesforce's interface layer that brings its CRM data, workflows, and governance into Claude, Slack, and Agentforce.
Source record checked 23 Sept 2026
Amazon
Amazon's preview integration paths for Alexa+, including smart-home tooling, MCP services and Wallet.
Source record checked 16 Oct 2026
Google DeepMind
Google's Gemini-powered evolutionary code-optimization agent, made generally available on Google Cloud on July 10, 2026.
Source record checked 23 Sept 2026
Arize
Arize AX made sessions a first-class object and expanded Agent-as-a-Judge evaluation to every pricing tier.
Source record checked 23 Sept 2026
Google open-sourced AX, an Apache-2.0 Kubernetes-style declarative runtime for orchestrating autonomous AI agents at scale.
Source record checked 23 Sept 2026
Meta
Meta's non-invasive brain-to-text research model that decodes typed sentences from MEG brain scans, aimed at communication for people with speech or motor disorders.
This entry is marked for review.
Source record checked 23 Sept 2026
OpenAI
A ChatGPT Work variant for investment banking and equity research, combining GPT-6 Astra with built-in premium financial data providers.
Source record checked 23 Sept 2026
OpenAI
ChatGPT's agent mode for multi-step work across connected apps, with synced tasks on desktop, web, and mobile.
Source record checked 1 Oct 2026
Anthropic
Anthropic's agentic coding CLI/desktop app shipped weekly updates through August-October 2026, including /design, claude plugin eval, default auto mode, and Mods.
Source record checked 6 Oct 2026
Anthropic
Anthropic's latest GA frontier model for coding and knowledge work, with 75% cheaper cache reads than Fable 5.
Source record checked 23 Sept 2026
Anthropic
Anthropic's near-frontier model delivering close to top-tier performance at roughly half the price.
This entry is marked for review.
Source record checked 23 Sept 2026
Cloudflare
Cloudflare's open-source workspace for agents, connected company systems and modifiable apps.
Source record checked 16 Oct 2026
Cobalt (cobaltcode.ai)
Gives coding agents (Codex, Cursor, OpenCode, Pi) a dedicated persistent VM per task with a live app preview and PR workflow.
This entry is marked for review.
Source record checked 23 Sept 2026
Cerebras
Cerebras' CS-4 wafer-scale AI system, the first built on its reusable Nexus rack-scale platform, unveiled at Supernova 2026.
Source record checked 23 Sept 2026
Cursor
Cursor's coordinator-and-subagent system that runs long, multi-week engineering work like migrations to completion via a single /goal.
Source record checked 23 Oct 2026
DeepSeek
DeepSeek's flagship reasoning model reaches general availability with a 1M-token context and expanded agentic tool use.
Source record checked 23 Sept 2026
DeepSeek
DeepSeek's 552B-parameter MoE model with native image+text input, a 1M-token context window, and MIT license.
Source record checked 18 Oct 2026
DigitalOcean
DigitalOcean's managed service combining isolated agent execution, governed tool access, and serverless inference in one cloud offering.
Source record checked 23 Sept 2026
Equinix
Equinix announced a distributed AI inference program with NVIDIA and Together AI supporting 200+ open models, availability Q1 2027.
This entry is marked for review.
Source record checked 23 Sept 2026
CrowdStrike
CrowdStrike's AI Detection and Response product that discovers, monitors, and enforces policy on agentic AI running inside an organization.
Source record checked 8 Nov 2026
Black Forest Labs
Multimodal flow model unifying video, audio and (soon) image generation, priced per second, with a robotics offshoot running on Audi production lines.
Source record checked 23 Sept 2026
Google DeepMind
Google's security-focused Gemini variant for vulnerability work within CodeMender's restricted pilot.
Source record checked 16 Oct 2026
Google DeepMind
Google's fastest, cheapest Gemini 3 model for high-throughput agentic search and document processing.
Source record checked 23 Sept 2026
Google DeepMind
Google's 1M-context workhorse Gemini model, faster and cheaper than 3.5 Flash, launched alongside 3.5 Flash-Lite and 3.5 Flash Cyber.
Source record checked 23 Sept 2026
Google DeepMind
Google's most intelligent workhorse Gemini model yet for coding and agents, with introductory $0.75/M input pricing.
Source record checked 23 Sept 2026
Google DeepMind
Google's cost-efficient real-time voice model for natural conversation with live visual context and background tasks.
Source record checked 23 Sept 2026
Google DeepMind
Google's live voice model for conversations that require multi-step reasoning and background tool use.
Source record checked 16 Oct 2026
A NotebookLM-style notebook feature built directly into the Gemini app, rolling out to Workspace and personal accounts.
Source record checked 23 Sept 2026
Google DeepMind
Google DeepMind's embodied-reasoning model that acts as a high-level planning brain for robots, with multi-robot collaboration.
Source record checked 23 Sept 2026
GitHub Copilot
GitHub Copilot Fusion, GA since Oct 28, 2026, routes each coding task across multiple models; successor to the HydraFusion preview.
Source record checked 29 Oct 2026
Zhipu AI
Zhipu's flagship coding/reasoning model with a 1M-token context, open-weighted about two weeks after launch.
This entry is marked for review.
Source record checked 10 Oct 2026
Google Cloud
An installable plugin that equips AI coding agents (Claude Code, Codex CLI, Antigravity) with Google Cloud skills, guardrails, and docs access.
Source record checked 23 Sept 2026
OpenAI
OpenAI's flagship model for computer/browser use, coding, and cybersecurity, rolling out via ChatGPT and API.
Source record checked 1 Oct 2026
OpenAI
OpenAI's most efficient GPT-6 model for high-volume, low-cost tasks like summarizing, extracting, and quick Q&A.
Source record checked 10 Oct 2026
OpenAI
GPT-6.1 Sol updates OpenAI's Sol line with near-Astra performance, lower cache-read pricing, and Multi-agent beta.
Source record checked 1 Oct 2026
OpenAI
OpenAI's mid-tier Terra model, relaunched as GPT-6.1 Terra on Oct 26, 2026 at $1 input and $5 output per million tokens.
Source record checked 27 Oct 2026
OpenAI
OpenAI's full-duplex voice model that listens and speaks simultaneously, now available in the API at $0.05/minute.
This entry is marked for review.
Source record checked 18 Oct 2026
OpenAI
OpenAI's frontier reasoning model built for biology, drug discovery, and translational medicine research, named after Rosalind Franklin.
Source record checked 2 Nov 2026
IBM
IBM's open-weight enterprise reasoning model family in 3B/8B/30B sizes with native chain-of-thought and tool calling.
Source record checked 23 Sept 2026
xAI
xAI's latest flagship reasoning model, unveiled via livestream, with native multi-agent "Colony" orchestration and a 1M-token multimodal context window.
Source record checked 14 Oct 2026
xAI
xAI's autonomous 'AI teammate' agents that run persistently in the cloud and now ship with more SuperGrok and Cursor plans.
Source record checked 23 Sept 2026
xAI
xAI's open-source terminal coding agent that reads and edits codebases, runs commands, and spawns up to eight subagents.
Source record checked 23 Sept 2026
Harvey
Legal AI platform ships Agentic Vault Search, Horizon Scanning, and Harvey Academy; raises $550M at a $15.5B valuation.
Source record checked 23 Sept 2026
Hebbia
Hebbia launches Max, a dealmaking AI agent, and Matrix 2.0, an agent that takes actions across financial workflows.
Source record checked 23 Sept 2026
Thinking Machines Lab
Thinking Machines Lab's open-weight multimodal MoE model with a 1M-token context and native text, image, and audio input.
Source record checked 23 Sept 2026
Moonshot AI
Moonshot AI's preview multimodal upgrade to its Kimi K2 open-weight agentic model line, released September 2026.
This entry is marked for review.
Source record checked 23 Sept 2026
Moonshot AI
Moonshot's 2.8-trillion-parameter model, the largest open weights ever released, under a custom revenue-gated license.
Source record checked 8 Nov 2026
Langfuse
Langfuse v4 went GA for self-hosted deployments with 10x faster dashboards and new monitors/code evaluators.
Source record checked 23 Sept 2026
Convai Innovations
An open-weight, self-hostable typed-decision model positioned as the open-source alternative to TypeSafe AI's Jev.
Source record checked 23 Sept 2026
Mistral AI
Mistral's open-weight Lean 4 theorem-proving model that solved 587 of 672 PutnamBench problems and saturated miniF2F.
Source record checked 23 Sept 2026
Hugging Face
Hugging Face's open-source robotics toolkit adds world-model policies, reward models, and simulation benchmarks to close the robot learning loop.
Source record checked 23 Sept 2026
Liquid AI
Liquid AI's open-weight LFM2.5-Encoder-230M and -350M models, bidirectional encoders tuned for fast long-context CPU inference.
Source record checked 23 Sept 2026
Anthropic
Vetted-access program giving life-science researchers Claude Mythos, Opus, and Sonnet with more permissive biology safeguards.
Source record checked 23 Sept 2026
Microsoft
Microsoft merged Copilot and Microsoft 365 Copilot into one unified app, rolling out to Windows and macOS from mid-September 2026.
Source record checked 23 Sept 2026
Xiaomi
Xiaomi's open-weight omnimodal flagship, the top-scoring open-weight model on Artificial Analysis at launch.
Source record checked 23 Sept 2026
Mistral AI
Mistral's production AI platform for observability, agent orchestration, and governed prompt/skill management across enterprise deployments.
Source record checked 23 Sept 2026
Mistral AI
Mistral's document-parsing model that outputs bounding boxes, block classification, and confidence scores across 170 languages for enterprise RAG pipelines.
Source record checked 23 Sept 2026
Modal
Modal's OpenAI-compatible, token-billed inference endpoint, launched with day-zero Kimi K3 support.
Source record checked 23 Sept 2026
Model Context Protocol
MCP's anniversary spec revision makes the protocol stateless, cacheable, and header-routable, with hardened OAuth and a formal async Tasks extension.
Source record checked 23 Sept 2026
Anthropic
Anthropic's specification letting AI agents safely operate physical lab and manufacturing hardware, extending agent tool-use to the physical world.
Source record checked 23 Sept 2026
Meta
Meta's personal AI agent that completes real tasks like booking and shopping on a dedicated secure cloud VM, not just a chatbot.
Source record checked 25 Sept 2026
Meta
Meta's open-weight 30B-parameter agentic model, distilled from Muse Spark, that runs locally on consumer hardware with tool use and multimodal input.
Source record checked 23 Sept 2026
Meta
Meta Superintelligence Labs' first image-generation model, which invokes search and coding tools to self-refine outputs rather than mapping prompts directly to pixels.
Source record checked 23 Sept 2026
Meta
Meta's hosted frontier reasoning model with a 1M-token context window, built for agentic and computer-use workflows, launched with the Meta Model API.
Source record checked 23 Sept 2026
Meta
Meta Superintelligence Labs' text-to-video model with native audio, built on the same pretraining base as Muse Image and previewed alongside it.
This entry is marked for review.
Source record checked 23 Sept 2026
NVIDIA
NVIDIA's open-weight model, downloadable and modifiable without a license fee.
This entry is marked for review.
Source record checked 23 Sept 2026
Cohere
Cohere Labs' 2.4B-parameter open-weight vision-language model built for native-resolution document understanding on edge hardware.
Source record checked 23 Sept 2026
Notion
Notion 3.7's shared skill library and agent workflow updates for teams.
Source record checked 16 Oct 2026
Ollama
Ollama v0.34.0 brings official ChatGPT Desktop integration on macOS, letting local models run inside ChatGPT's app.
Source record checked 23 Sept 2026
OpenAI
OpenAI's managed enterprise platform for deploying governed voice and chat agents into high-volume, high-stakes production workflows.
Source record checked 23 Sept 2026
Opera
Opera's agentic AI browser dropped its paid-only model to become a free download with an MCP connector and CLI.
Source record checked 23 Sept 2026
Perplexity
Perplexity's agentic AI browser, now on all major platforms, with local-desktop context via Personal Computer for Mac.
Source record checked 23 Sept 2026
Pinecone
Pinecone's knowledge-orchestration layer that compiles enterprise data into cited artifacts agents can query directly.
Source record checked 23 Sept 2026
Qwen
Alibaba's flagship 2.4T-parameter model, the first open-weight Max-tier Qwen release.
This entry is marked for review.
Source record checked 23 Sept 2026
Qwen
Alibaba's omnimodal model taking text, image, audio, and video in one request with a 1M-token context window.
Source record checked 23 Sept 2026
Replit
Replit added automatic model routing, enterprise governance, pen-testing, and GA Databricks Lakebase support across August-September 2026.
This entry is marked for review.
Source record checked 23 Sept 2026
Mistral AI
Mistral's 8B robotics navigation model that steers wheeled, legged, or flying robots using a single RGB camera and plain-language instructions.
Source record checked 23 Sept 2026
CrowdStrike
A CrowdStrike-NVIDIA agentic system built on Nemotron models for autonomous cyber defense, including AI-vs-AI red/blue teaming.
This entry is marked for review.
Source record checked 23 Sept 2026
ElevenLabs
ElevenLabs' batch speech-recognition model specialized for medical and clinical audio, now generally available.
Source record checked 23 Sept 2026
Mistral AI
Mistral's 3B open-weight, policy-adaptive safety classifier for text and image moderation.
Source record checked 23 Sept 2026
Slack
Slack workspace feature turning channels into multiplayer environments for AI coding agents like Claude Code, Devin, and Copilot.
Source record checked 23 Sept 2026
Zendesk
Zendesk's industry- and business-specific AI agents, launching in commerce, that automate up to 80% of workflows.
Source record checked 23 Sept 2026
OpenAI
OpenAI's in-ChatGPT ad format that opens a labeled, direct conversation with a brand's own AI agent instead of linking out.
Source record checked 23 Sept 2026
Suno
Suno's v6 music-model family, with flagship, exploratory and smaller variants.
Source record checked 16 Oct 2026
Zephyr
A desktop workspace, from the team behind Module Federation, where teams and reusable AI 'Specialists' collaborate in shared channels and build workflow mini-apps.
This entry is marked for review.
Source record checked 23 Sept 2026
Thomson Reuters
Thomson Reuters' first in-house proprietary large language model, trained on its Westlaw and Practical Law content for professional-grade accuracy.
Source record checked 23 Sept 2026
Together AI
Together AI's hosted fine-tuning platform adds a no-code UI, Direct Preference Optimization, continued training, and Expert LoRA for MoE models.
Source record checked 23 Sept 2026
Hugging Face
A September 2026 Transformers release adding model architectures and runtime changes.
Source record checked 16 Oct 2026
vLLM Project
Open-source LLM inference/serving engine; v0.31.0 (October 13, 2026) adds GA disaggregated prefill/decode serving and day-0 support for Grok 5 and Mistral Large 4.
Source record checked 14 Oct 2026
Mistral AI
Mistral's 24B open-weight speech-to-text model for transcription, translation, and audio understanding.
This entry is marked for review.
Source record checked 23 Sept 2026
Visual Studio Code
VS Code 1.138 lets AI coding agents run inside a project's local Dev Container, keeping agent work aligned to the project's own toolchain and dependencies.
Source record checked 23 Sept 2026
Cloudflare
Cloudflare added six new open-weight text models to Workers AI, including DeepSeek-V4 and Kimi K2.7-code.
Source record checked 23 Sept 2026
Huawei
Huawei's planned Ascend accelerator, with availability announced for the first quarter of 2027.
Source record checked 16 Oct 2026
TestMu AI
A Kane CLI extension that turns product requirement documents into permanently-linked, evidence-backed test coverage.
Source record checked 22 Sept 2026
Egnyte
Egnyte's business-context foundation that maps content, people, projects, and systems so AI agents get accurate context without redundant processing.
Source record checked 22 Sept 2026
CorelDRAW
CorelDRAW's September 2026 update adding AI text-to-vector generation and enhanced PowerTRACE bitmap-to-vector tracing.
Source record checked 22 Sept 2026
Factory
AI software-development platform whose 'Droid' agents plan, code, test, and document software; raised $200M at a $5B valuation in September 2026.
Source record checked 22 Sept 2026
xAI
xAI's speech-to-text model with diarization, timestamps, and smart turn detection for streaming or batch transcription.
Source record checked 22 Sept 2026
TypeSafe AI
TypeSafe AI's first System One model: a non-LLM decision engine returning typed, calibrated choices instead of generated text.
Source record checked 26 Oct 2026
CopilotKit
MIT-licensed platform for self-hosted AI coworkers, each with its own sandboxed browser, files, and policy-checked tools.
Source record checked 22 Sept 2026
Unbiased
Unbiased's blended model that runs several frontier and open-source models per request and returns the best synthesized answer.
Source record checked 22 Sept 2026
Qwen
Alibaba's Qwen team open-weight 7B image model unifying generation and editing with native transparent-PNG output, research-license only.
Source record checked 15 Oct 2026
ReleasePad
An MCP server that lets AI coding agents draft, publish, and measure product changelogs, with every post held for human approval.
Source record checked 22 Sept 2026
Cloudflare
Cloudflare's open-source coding-agent skill for multi-phase security audits producing independently verified, machine-readable findings.
Source record checked 22 Sept 2026
