Sarvam AI · Sarvam
Sarvam 105B
Sarvam's 106B/10.3B-active open-weights MoE reasoning model from India.
- AA Intelligence
- 8.8#108 of 142
- Price
- Self-host
- Output speed
- —
- Context window
- 128K tokens
Composite indexes
- Artificial Analysis Intelligence Index#108 of 1429
Leader: Claude Opus 5.5 at 58
Agents and tool use
- τ²-Bench#47 of 8846.8%
Leader: GLM-5.2 at 99.1%
Reasoning and knowledge
- GPQA Diamond#93 of 14273.8%
Leader: GPT-6 Astra at 96.1%
- Humanity’s Last Exam#96 of 14711%
Leader: Claude Opus 5.5 at 61.4%
- CritPt#136 of 1390%
Leader: GPT-5.6 Sol at 32.3%
Vision and long context
- AA Long Context Reasoning#138 of 1390%
Leader: Kimi K3 at 88.7%
Other published results
| Result | Score | Reported by | Source |
|---|---|---|---|
| Terminal-Bench Hard (AA) · high effort | 1.5% | independent | Artificial Analysis: Sarvam 105B (high) ↗ |
| IFBench · high effort | 34.4% | independent | Artificial Analysis: Sarvam 105B (high) ↗ |
| AA-Omniscience Index · high effort | -59.4 index | independent | Artificial Analysis: Sarvam 105B (high) ↗ |