By No Code MBA · Facts checked
- Provider
- Perplexity
- Announcement date
- Status at last check
- Generally available
- Model family
- pplx-decider
- Preceding model
- pplx-decider-v1-27b
- Capabilities
- Text, Image
What changed
pplx-decider-v1.1-27b is an updated open-weight decision model built on Qwen3.8-27B (about 26B parameters) with a separate decision head that returns a probability for each candidate answer instead of generating text. Perplexity's model card says full-attention layers now use noncausal attention, training data grew, and the image size cap is configurable rather than fixed at 512x512. Perplexity reports its Decision Index score rose from 56.4 to 61.56, and says it costs half as much as v1.
Access and availability
Weights are on Hugging Face at perplexity-ai/pplx-decider-v1.1-27b under Apache 2.0 and run with Perplexity's provided Python code.
Official access documentationPricing
$0.02 per million input tokens
Official pricingWhat you can build
Evaluate it on a small set of representative classification or routing examples, comparing decision probabilities with your current approach and checking uncertain cases with human review.
Limitations
Text and images for one decision must fit within 8,192 tokens, and oversized inputs are rejected. It needs a CUDA GPU with about 49 GiB for weights, noncausal attention must be enabled to match reported results, and it has no full language-model head for open-ended text.