Model profile
Sarvam 105B
Evidence coverage
11 of 323 tracked benchmarks are published. 0 are verified and 11 provisional. 5 of 8 categories are measured.
- Published / tracked
- 11 / 323
- Verified
- 0
- Provisional
- 11
- Categories with evidence
- 5 / 8
Evidence by category
- Agentic1 benchmarkReported
- Coding1 benchmarkReported
- Reasoning2 benchmarksReported
- Knowledge6 benchmarksReported
- Math0 benchmarksNot measured
- Multilingual0 benchmarksNot measured
- Multimodal0 benchmarksNot measured
- Inst. Following1 benchmarkReported
Sarvam 105B ranks #157 out of 200 models on the public leaderboard with an overall score of 42.97/100. It does not yet have enough sourced coverage for BenchLM's verified leaderboard. While not a frontier model, it offers specific advantages depending on the use case.
Sarvam 105B is a open weight model with a 128K token context window. It uses explicit chain-of-thought reasoning, which typically improves performance on math and complex reasoning tasks at the cost of higher latency and token usage.
This profile currently has 11 of 323 tracked benchmarks. BenchLM only exposes non-generated benchmark rows publicly, so missing categories stay blank until a sourced evaluation is available.
Peer position
Exact provisional scores and ranks for the closest listed peers. A score can appear before a model clears the evidence threshold for a rank, so equal scores can have different rank states.
Range 42.58–43.87
- Ling 2.6 FlashInclusionAICompare#15443.87Ling 2.6 Flash is #154 with a score of 43.87.
- Gemma 4 E4BGoogleCompare#15543.2Gemma 4 E4B is #155 with a score of 43.2.
- Mistral Medium 3MistralCompare#15643.2Mistral Medium 3 is #156 with a score of 43.2.
- Sarvam 105BCurrent modelSarvam#15742.97Sarvam 105B is #157 with a score of 42.97.
- Claude 4 SonnetAnthropicCompare#15842.79Claude 4 Sonnet is #158 with a score of 42.79.
- GPT-OSS 20BOpenAICompare#15942.74GPT-OSS 20B is #159 with a score of 42.74.
- DeepSeek R1 Distill Qwen 32BDeepSeekCompare#16042.58DeepSeek R1 Distill Qwen 32B is #160 with a score of 42.58.
Category percentile
More
Relative position among models eligible for each sourced category. A higher percentile means a stronger position within that category's ranked cohort; 100 is highest.
Category evidence
Scores and ranks appear only where this model has published benchmark evidence. Categories without displayable source records remain not measured.
| Category | Score | Rank | Percentile | Weight | Benchmarks | Evidence |
|---|---|---|---|---|---|---|
| AgenticRank Not rankedWeight 22%1 benchmarkReported | 49.5 | Not ranked | Not available | 22% | 1 benchmark | Reported |
| CodingRank Not rankedWeight 20%1 benchmarkReported | 45.0 | Not ranked | Not available | 20% | 1 benchmark | Reported |
| ReasoningWeight 17%2 benchmarksReported | Score pending | Not ranked | Not available | 17% | 2 benchmarks | Reported |
| KnowledgeRank Not rankedWeight 12%6 benchmarksReported | 39.4 | Not ranked | Not available | 12% | 6 benchmarks | Reported |
| MathWeight 5%0 benchmarksNot measured | Not measured | Not ranked | Not available | 5% | 0 benchmarks | Not measured |
| MultilingualWeight 7%0 benchmarksNot measured | Not measured | Not ranked | Not available | 7% | 0 benchmarks | Not measured |
| MultimodalWeight 12%0 benchmarksNot measured | Not measured | Not ranked | Not available | 12% | 0 benchmarks | Not measured |
| Inst. FollowingRank Not rankedWeight 5%1 benchmarkReported | 84.8 | Not ranked | Not available | 5% | 1 benchmark | Reported |
Benchmark Details
Rows below have a displayable published verification record. Each source link and provenance note remains in the page HTML while its category is closed. Source-unverified manual rows and generated rows stay hidden.
Agentic1 benchmark
τ²-Bench Tool-Agent-User Evaluation
Coding1 benchmark
Artificial Analysis SciCode
Reasoning2 benchmarks
Artificial Analysis Long Context Reasoning
Critical Physics Tasks
Knowledge6 benchmarks
Artificial Analysis GPQA Diamond
Artificial Analysis Humanity's Last Exam
Artificial Analysis Omniscience Index
Artificial Analysis Omniscience Accuracy
Artificial Analysis Omniscience Hallucination Rate
Inst. Following1 benchmark
Artificial Analysis IFBench
Frequently Asked Questions
How does Sarvam 105B perform overall in AI benchmarks?
Sarvam 105B has 11 published benchmark scores on BenchLM, but it does not yet have enough non-generated coverage to receive a global overall rank.
Is Sarvam 105B good for knowledge and understanding?
Sarvam 105B has visible benchmark coverage in knowledge and understanding, but BenchLM does not currently assign it a global category rank there.
Is Sarvam 105B good for coding and programming?
Sarvam 105B has visible benchmark coverage in coding and programming, but BenchLM does not currently assign it a global category rank there.
Is Sarvam 105B good for reasoning and logic?
Sarvam 105B has visible benchmark coverage in reasoning and logic, but BenchLM does not currently assign it a global category rank there.
Is Sarvam 105B good for agentic tool use and computer tasks?
Sarvam 105B has visible benchmark coverage in agentic tool use and computer tasks, but BenchLM does not currently assign it a global category rank there.
Is Sarvam 105B good for instruction following?
Sarvam 105B has visible benchmark coverage in instruction following, but BenchLM does not currently assign it a global category rank there.
Is Sarvam 105B open source?
Yes, Sarvam 105B is an open weight model created by Sarvam, meaning it can be downloaded and run locally or fine-tuned for specific use cases.
Does Sarvam 105B have full benchmark coverage on BenchLM?
Not yet. Sarvam 105B currently has 11 published benchmark scores out of the 323 benchmarks BenchLM tracks. BenchLM only exposes non-generated public benchmark rows, so missing categories stay blank until a sourced evaluation is available.
What is the context window size of Sarvam 105B?
Sarvam 105B has a published context window of 128K, which determines how much text it can process in a single interaction.
Related Resources
Choose with this week’s evidence
Join 2,000+ readers for ranking moves, new releases, pricing changes, and the evidence behind them.
Free. One email per week.