pplx-decider v1.1

Perplexity's updated open model for picking the right label or route for a request, at half the price of v1, for developers using the Decisions API.

A 27B-parameter model trained to choose among options you define and return a probability for each, rather than write free text.
Takes text and image input, with a 250K token context window.
Costs $0.02 per million input tokens through the Perplexity Decisions API, half the price of v1.
Open weights under the Apache 2.0 license on Hugging Face. Running it yourself takes about 49GB of GPU memory.
Built on a Qwen3.8-27B base. Perplexity says it scores highest on Hugging Face's new Decision Index benchmark.