Perplexity Decider v1.1 (27B)
Perplexity Decider v1.1 (27B) by Perplexity is a decision model: it answers typed questions with a probability for each allowed answer instead of generating text. It scores 62.8 on the Jev Decision Index (rank 3). Hosted, it costs about $0.0040 per 1,000 decisions.
Pricing
| Provider | Input / 1M | Output / 1M |
|---|---|---|
| perplexity | $0.02 | $0.00 |
| openrouter | $0.02 | $0.00 |
Cost per 1,000 decisions assumes 200 input tokens per decision.
Specifications
- Provider
- Perplexity
- Parameters
- 28B
- Released
- Oct 2026
- Open weights
- Yes
- Question types
- Choice (one of N options), Score (ordered rubric), Yes / no
- Mechanism
- Decoder LM with a decision readout
- Backbone
- Qwen/Qwen3.8-27B
- Jev-compatible request
- Yes
- Runs on CPU
- No
- License
- apache-2.0
- Hugging Face
- perplexity-ai/pplx-decider-v1.1-27b
Benchmarks
- Jev Decision IndexECE 0.071
- 62.8
- Jev Decision Index — VisionECE 0.044
- 70.6
Jev Decision Index: public 62.3, ECE 0.071, Brier 0.335. Lower ECE means its probabilities can be taken at face value.
Speed & latency
Compare Perplexity Decider v1.1 (27B) with…
FAQ
How much does Perplexity Decider v1.1 (27B) cost?
Hosted, Perplexity Decider v1.1 (27B) costs $0.02 per 1M input tokens, about $0.0040 per 1,000 decisions at 200 input tokens each.
What questions can Perplexity Decider v1.1 (27B) answer?
Perplexity Decider v1.1 (27B) answers choice (one of n options), score (ordered rubric), yes / no questions, returning a probability for each allowed answer.
Is Perplexity Decider v1.1 (27B) well calibrated?
On the Jev Decision Index its expected calibration error (ECE) is 0.071; below about 0.05, a stated 80% confidence is right roughly 80% of the time.
Is Perplexity Decider v1.1 (27B) open weight, and can it run on a CPU?
Yes, Perplexity Decider v1.1 (27B) has open weights. It needs a GPU to run at useful speed.
Data as of 2026-10-14. Source: Metatext decision-model catalogue, 2026-10-14 (OpenRouter, Jev Decision Index, Benchmark Heaven JevBench, Hugging Face). Pricing is per 1M tokens (USD); confirm with the provider before production use.