Sarvam Prism: open-weight decision model that answers in 44 milliseconds
Sarvam AI launched Prism, a System 1 decision model built on Sarvam 105B that returns a probability score for each available option in a single forward pass, averaging 71.8 across three benchmarks at 44 ms on an NVIDIA B200 GPU, with weights released under Apache 2.0.
Why it matters
Decisions as single-pass scoring instead of token generation: sub-100 ms structured judgments with calibrated confidence, in open weights that enterprises can fine-tune and run on their own infrastructure.
What you could build with it
A startup building customer-support automation could route every incoming ticket through Prism to pick the right queue, reply template, or escalation path in milliseconds, replacing a slow generative call for each routine classification decision.
Does it hold up?
Too early to judge: only vendor benchmarks have appeared so far (71.8 average, 44 ms on a B200); independent deployments and the promised benchmark release are still to come.
Built with Sarvam Prism
- Sarvam Prism: System 1 decision-making at 44 ms, tested against 25 modelsfortuneindia · Coverage of Prism's System 1 architecture, its 71.8 average score, and its calibrated-confidence results on Indian-language tasks.
- Sarvam AI launches Prism model for enterprise decision-makingthehindubusinessline · Business Line's report on the launch, the Apache 2.0 weight release, and the enterprise use cases Sarvam is targeting.
- Sarvam Prism scores 82.5 on an internal Indian-language benchmarkpunjabkesari · ANI report noting Prism's Indian-language benchmark score against Perplexity Decider and Jev, plus the open-weight release.
Learn more
First spotted on article: source.
More AI models
VoiceStudio: open-source, fully-local ElevenLabs alternative for voice cloning, dubbing, and audiobooksdebpalash/VoiceStudio is a desktop app that runs voice cloning, voice design, video dubbing, dictation, transcription…models · JEV 0.75Tencent Hunyuan Hy3: open 295B MoE with 21B active under Apache 2.0Tencent open-sourced Hy3, a 295B-parameter mixture-of-experts model with only 21B parameters active per token, under…models · JEV 0.73NVIDIA Isaac Lab 3.0 GA + Isaac OS open runtime + Agile ONE SN humanoidMajor Isaac platform expansion at GTC Berlin Oct 21: Isaac Lab 3.0 is generally available as the open-source robot…models · JEV 0.73NVIDIA Nemotron 3.5 Retriever collection: Embed 8B/1B, Rerank 8B, Parse 631MNVIDIA's Oct 21 collection drop for agentic retrieval: Nemotron-3.5-Embed in 8B BF16, 1B BF16 and 1B NVFP4…models · JEV 0.72Kimi K3.1 Pro: Moonshot's 2.8T open-weight flagship officially launchesMoonshot AI officially launched Kimi K3.1 Pro on Oct 23, a mixture-of-experts model with 2.8 trillion total / 104…models · JEV 0.72NVIDIA Cosmos 3.5 world models: now openly availableNVIDIA's Cosmos 3.5 family of open world foundation models for robots, autonomous vehicles and vision agents is now…models · JEV 0.71
Get the week's best AI launches, plus 3 ideas worth building
One email every Saturday. Ranked by traction, not hype. Free.