Sarvam releases Prism, an open-weights decision model scoring 71.8 at 44ms latency
itsOmSarraf_ · x · 2026-11-03
Indian AI company Sarvam released Sarvam Prism, a "System One" decision model built on Sarvam 105B for fast, accurate decisions in enterprise workflows.
- Scores 71.8 with a 44ms median response time, claimed best average among decision models responding under 100ms
- Weights are on Hugging Face (sarvamai/prism) under Apache 2.0, covering English plus Hindi, Bengali, Tamil and other Indian languages
- Tagged as text-classification/decision-model with calibration and custom code
Model release shared via a third-party retweet.
Related event: India's Sarvam Open-Sources Prism, a 44ms Decision Model(3 posts)→
More from Models
- JFPuget's team takes 3rd on both Kaggle ARC-AGI competitions ending the same day — JFPuget · 2026-11-04
- Tencent's WeChat Vision team open-sources WeVisOne-VL, a unified 2B/4B perception VLM — bdsqlsz · 2026-11-04
- DeepMind-Stanford paper: a red heart in your prompt flips code from 0% to 100% recursive — rohanpaul_ai · 2026-11-04
- A Red Heart Can Steer Code: DeepMind-Stanford Paper Probes Visual Meaning in VLMs — rohanpaul_ai · 2026-11-04
- DeepSeek open-sources DeepInfer inference engine delivering 26,000+ tokens/sec per GPU — bycloud · 2026-11-04
- DeepMind RL-trained Gemini doubles correct psychiatric differential diagnoses in blind study — SRSchmidgall · 2026-11-04