Perplexity AI releases decision model pplx-decider-v1-27b fine‑tuned from Qwen3.8-27B
Perplexity AI has open‑sourced a new decision‑making model called pplx-decider-v1-27b. The model is a fine‑tune of the Qwen3.8‑27B base and was evaluated on eleven public benchmarks, with the best scores highlighted in the release notes. Accuracy figures are reported through the Perplexity API, though the article does not quote exact percentages.
Key points
- pplx-decider-v1-27b is a decision model fine‑tuned from Qwen3.8‑27B and evaluated on eleven benchmarks
- The model needs about 49 GiB of GPU memory and runs on Python 3.12+ with CUDA
- Weights are hosted on Hugging Face and can be run via a short uv‑based inference script
The model runs on Python 3.12+ with a CUDA‑enabled GPU and requires roughly 49 GiB of VRAM for the weight files plus additional working memory. Users can download the weights from Hugging Face and run inference via a simple script or the provided uv commands. Example code shows how to classify a support ticket and how to request a yes/no probability for urgency, with optional image input.
Perplexity AI positions the decider as a lightweight, open‑weight alternative for developers needing calibrated choice predictions without building a full‑scale LLM pipeline. The release includes a ready‑to‑run inference example and clear installation steps, making it immediately usable for downstream applications such as ticket routing or urgency detection.
Model page: pplx-decider-v1-27b →
The story so far
3 episodes →- Perplexity AI releases decision model pplx-decider-v1-27b fine‑tuned from Qwen3.8-27Bthis story
Perplexity Decider 27B: Open weights decision model fine tune of Qwen3.8 27B
huggingface.co · 2 October 2026
Loading the full article…
This text was published by huggingface.co. It is reproduced here with attribution so you can read it in full; the rights remain with the publisher. Read it at the source ↗
Coverage and discussion
1source- Reddit discussionreddit.com
The headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us. Published by Martin K., who runs Digest AI and handles corrections.
More in Generative AI & Models
All →- OpenAI releases GPT-6.1 Sol, says it nearly matches Astra on agentic coding at one-fifth the price · 37 src
- Author finds 'just right' prompt range speeds GPT-6 Luna High · 1 src
- Google rolls out Gemini 4 Argon, its most advanced AI model · 16 src
- Google launches Gemini 4 Argon with 1M token output, claims lead on DeepSWE and cybersecurity benchmarks · 19 src
- Tavus' Griffin AI avatar fools 48% of users in one-minute call · 1 src
Comments
via GitHub Discussions