pplx-decider-v1.2-27b

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.

Perplexity · Open-weight/community models · model

Published in Feed

Overview

Perplexity shipped pplx-decider-v1.1-27b on October 6, 2026, an update to its open-weight multimodal decision model built on a Qwen3.8-27B backbone. It answers typed Choice, Score, and Null questions about text or images with calibrated probabilities through Perplexity's Decisions API, replacing v1 as the default. It scores highest on the new Hugging Face Decision Index 0.3 benchmark and costs half as much to run as v1, at $0.02 per million input tokens, with weights published ungated on Hugging Face under Apache 2.0 and availability on OpenRouter the following day. On October 16-17, 2026 Perplexity followed with pplx-decider-v1.2 (27B) and pplx-decider-flash-v1.2 (4B), open-weight multimodal decision models with the same API that Perplexity says run up to 5x faster, with Flash billed as its top 4B model on its own Decision Index. Roundups list prices of $0.02 and $0.01 per million tokens, with availability on Hugging Face, the Perplexity API, and OpenRouter; OpenRouter lists Flash with a 262K context window.

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

  • 250K-token context window; accepts text and image input
  • Answers typed Choice, Score, and Null questions with calibrated probabilities instead of free text
  • Built on a Qwen3.8-27B backbone, open weights on Hugging Face under Apache 2.0 (ungated)
  • Tops the Hugging Face Decision Index 0.3 benchmark
  • Half the input-token price of v1: $0.02 per million input tokens
  • v1.2 (Oct 16-17, 2026): 27B model plus a 4B pplx-decider-flash-v1.2, same API, vendor-claimed up to 5x faster inference

Use cases

  • Structured decision/classification steps inside agent pipelines via Perplexity's Decisions API
  • Low-cost routing, moderation, or scoring over text-plus-image inputs
  • Self-hosting the open weights for private decision-model inference

Pricing

$0.02 per million input tokens via Perplexity's Decisions API (half of v1's price); v1.2 Flash (4B) listed at $0.01 per million tokens; open weights free to self-host under Apache 2.0 (v1.1).

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

Dates and sources

Released
6 Oct 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