DeepSeek-V4.1-Flash

DeepSeek's 552B-parameter MoE model with native image+text input, a 1M-token context window, and MIT license.

DeepSeek · Open-weight/community models · model

Published in Feed

Overview

DeepSeek-V4.1-Flash is a 552-billion-parameter Mixture-of-Experts model (8B active on prefill, 16B on decode) released September 10, 2026, reading images and text natively with context up to 1 million tokens and up to 384,000 output tokens. It uses a Compressed Expert Dispatch (CED) architecture and is published on Hugging Face under the MIT license with an accompanying technical report. On October 17, 2026 DeepSeek shipped DeepSeek-V4.1-Flash-1017, a same-price update trained with reinforcement learning from self-feedback for end-to-end project work such as game creation and 3D modeling, optimized for DeepSeek Harness, Tencent CodeBuddy, and WorkBuddy. A third-party tracker reports Terminal-Bench 4.0 rising from 31.2 to 42.3; DeepSeek's own page shows charts without text scores. The API name stays deepseek-flash, and Search and Fetch endpoints became generally available at $0.75 per 1,000 calls, free through server-side tools until October 31, 2026 at 16:00 UTC.

This is a sourced directory entry. Features and capabilities are reported in the linked sources; this entry does not document independent Aymorphic testing.

Key features

  • 552B total parameters, 8B/16B active (prefill/decode)
  • 1M-token context, 384K max output
  • Native text + image input
  • Compressed Expert Dispatch (CED) architecture
  • MIT license
  • V4.1-Flash-1017 (Oct 17, 2026): RL-from-self-feedback update for end-to-end project tasks, same API name and token prices
  • Search and Fetch APIs at $0.75 per 1,000 calls each; free via server-side tools until Oct 31, 2026

Use cases

  • Cost-sensitive multimodal agentic tasks
  • High-volume API workloads (peak/off-peak pricing)

Pricing

MIT license, weights free; API peak $0.30/M input (cache miss), $0.006/M (cache hit), $1.20/M output; off-peak half those rates.

Pricing reflects the source record and may change. Confirm current terms with the provider.

Dates and sources

Released
10 Sept 2026
Record updated
18 Oct 2026
Record last checked
18 Oct 2026

Explore the Feed

Other entries filed under open-weight/community models; inclusion does not imply equivalent features.

  • Muse Spark Open

    Meta's open-weight 816B-parameter MoE model (42B active, 1M context) for coding and long-horizon agent work, released Oct 15, 2026.

  • Nemotron 3.5 Bolt

    NVIDIA's open 74B MoE model (9B active) packed into 24GB for local agents on RTX Spark; released Oct 15, 2026.

  • Jais 3

    Six open Arabic-focused language models from 900M to 375B parameters with 512K context, plus Jais Speech for Arabic recognition.

Browse the Feed