Sarvam AI Launches Prism, a Jev-Like Decision Model Built on Sarvam 105B
Bengaluru-based Sarvam AI released Prism, a LoRA fine-tune of its Sarvam 105B model designed for rapid decision-making tasks. The model processes input in a single forward pass and returns probability scores for available options, achieving 44-millisecond response times. Prism scored highest among 25 tested decision models with sub-100ms latency and runs on a single GPU with 120GB memory in FP8 precision. The open-source model on Hugging Face targets enterprise workflows including intent routing, customer support, and document intelligence.
Key points
- Single-pass decision model returning 44ms responses with 71.8 benchmark score
- Open-sourced under Apache 2.0 on Hugging Face for enterprise fine-tuning
- Runs on single GPU and available via Sarvam API
- Outperformed 24 other decision models in speed-accuracy tests
Summary and key points written by AI from the headline and a short excerpt, so they can be incomplete or wrong. Analytics India Magazine did the reporting; read the original for the full story.
Read the full story at Analytics India Magazine
