# Decider model sizes and sources

> Decider is Perplexity's decision model for text and images, with Apache-2.0 weights. V1.2 lists $0.02 per million input tokens on OpenRouter.

Updated 2026-11-04 (Australia/Adelaide)

By Perplexity · Released: 2026-10-06 · OPEN

Perplexity's decision model reads text and images and returns a probability for every answer. It ships as Apache-2.0 open weights and as a hosted API. V1.1 (27B) halved the price on 6 Oct; V1.2 and a 4B [Decider](<https://decisionmodels.ai/models/pplx-decider>) Flash followed on 20 Oct.

## Facts

- Latest: [Decider](<https://decisionmodels.ai/models/pplx-decider>) V1.2 (27B) and [Decider](<https://decisionmodels.ai/models/pplx-decider>) Flash V1.2 (4B)
- Price: $0.02 (V1.2) and $0.01 (Flash) per million input tokens on OpenRouter, output free (as of 22 Oct 2026) · [Compare prices](<https://decisionmodels.ai/pricing>)
- Licence: Apache-2.0

## Claims and sources

- **INDEPENDENT**: [Decider](<https://decisionmodels.ai/models/pplx-decider>) V1.1-27B was the most accurate of seven decision models (94.7%) on 799 unpublished yes/no questions about coding-agent work and destructive shell commands, and the only one to pass SILT's registered calibration rule. [Jev](<https://decisionmodels.ai/models/jev>)'s result was inconclusive. So were the results for OpenAI's [Decisions API](<https://decisionmodels.ai/models/openai-decisions>) and [Laya](<https://decisionmodels.ai/models/laya>) (11 Oct 2026) [source](<https://www.silt-seb.com/studies/decision-models>)

## Links

- [Source: V1.2 announcement](<https://community.perplexity.ai/t/pplx-decider-v1-2-is-the-highest-scoring-decision-model-on-claimbench/6392>)
- [Source: Hugging Face](<https://huggingface.co/perplexity-ai/pplx-decider-v1.2>)
- [Source: V1.1 announcement](<https://community.perplexity.ai/t/pplx-decider-v1-1-27b-our-updated-open-weights-multimodal-decision-model-is-now-available/6286>)

[JevBench scores and methodology](<https://decisionmodels.ai/jevbench>)

## Related projects

### Decision models forecasting Polymarket

[Decision models forecasting Polymarket](<https://decisionmodels.ai/projects#project-odds-arena>)

[Source: aaron-zhlg/odds-arena](<https://github.com/aaron-zhlg/odds-arena>)

Keeps the 50 busiest Polymarket questions forecast by several models, including [Jev](<https://decisionmodels.ai/models/jev>), Perplexity's [Decider](<https://decisionmodels.ai/models/pplx-decider>) and an OpenAI model, all given filtered web search results but never the price. When markets resolve, each is scored on a leaderboard by Brier score, log loss and accuracy.

- Creator: aaron-zhlg

- Stars: 0

- Licence: no licence

- Model: [Jev](<https://decisionmodels.ai/models/jev>), [Decider](<https://decisionmodels.ai/models/pplx-decider>)

- Tags: forecasting, benchmark

### Narrative audit: earnings calls against the numbers

[Narrative audit: earnings calls against the numbers](<https://decisionmodels.ai/projects#project-narrative-audit-jev>)

[Source: adityavsb/narrative-audit-jev](<https://github.com/adityavsb/narrative-audit-jev>)

A harness that checks claims from earnings-call transcripts and footnotes against tables in DuckDB, on 64 synthetic cases. Perplexity's pplx-decider-flash-v1.2 scored 85.94% at about 1.4 s a case; a rule baseline scored 68.75% but caught only 2 of 9 Q&A deflections, which the [decider](<https://decisionmodels.ai/models/pplx-decider>) caught 9 of 9. A chat model led on accuracy at 90.62%, with confidences stacked at 0.99 to 1.00. OpenRouter's chat endpoint rejected the [decider](<https://decisionmodels.ai/models/pplx-decider>) with HTTP 400, so it runs through the separate /api/alpha/decisions endpoint.

- Creator: adityavsb

- Stars: 0

- Licence: MIT

- Model: [Decider](<https://decisionmodels.ai/models/pplx-decider>)

- Tags: benchmark, finance