Perplexity releases Decider v1.3 open-weight decision models
Start here
Perplexity has released pplx-decider-v1.3 and pplx-decider-flash-v1.3, updated versions of its open-weight decision models that return per-option probabilities instead of generated text. The weights are published under Apache 2.0 on Hugging Face, and both models are also available through the Perplexity Decisions API and OpenRouter. The main addition is Locate, which extracts exact text spans with start and end character offsets for tasks such as named-entity recognition, PII detection and extractive question answering. Perplexity reports 91.6% mean accuracy on 63 query-classification tasks and an average of 69.0 on 17 out-of-domain Locate tasks; these figures are self-reported. API pricing is $0.02 per 1M input tokens for v1.3 and $0.01 for Flash, with output free.
Generated from reports · updated 16 min ago
Report timeline
Follow the reports to see the event from different sides.
- MarkTechPostPerplexity Releases Decider v1.3 Open-Weight Decision Models with Locate Span Extraction
Perplexity released pplx-decider-v1.3 and pplx-decider-flash-v1.3, open-weight decision models that return per-option probabilities rather than generated text, with weights under Apache 2.0 on Hugging Face and availability in the Perplexity Decisions API and OpenRouter. The main addition is Locate, which extracts exact text spans with start and end character offsets for NER, PII detection and extractive QA. Perplexity reports 91.6% mean accuracy on 63 query-classification tasks and 69.0 average on 17 out-of-domain Locate tasks, though all figures are self-reported; API pricing is $0.02 and $0.01 per 1M input tokens for v1.3 and Flash.
Heat for this event
Not enough continuous observations yet to draw a trend.