Sarvam AI model product
Sarvam 105B
Sarvam-105B is Sarvam AI's flagship open-source Mixture-of-Experts reasoning model built for complex reasoning, coding, and agentic workflows.
Updated Aug 12, 2026. Default version: Sarvam-105B
LLMBoard score44.6Sarvam-105B
Coverage100%14 benchmark families
Context window131.1KTokens
Official input priceN/AOfficial price unavailable
Sarvam 105B Specifications
Technical details for the model's default version.
- Version
- Sarvam-105B
- Released
- Mar 6, 2026
- Knowledge cutoff
- Unknown
- Parameters
- 105B
- Context window
- 131.1K
- Max output
- 131.1K
- Inputs
- text
- Outputs
- text
- Open weights
- Yes
- License
- Apache 2.0
Sarvam 105B Capability Profile
This profile uses the model's current scored version. Arena ratings and prices are shown separately.
Sarvam 105B Benchmark Results
Benchmark scores for Sarvam-105B.
| MATH-500 | 98.6% | 02 | 32 | 96.8% | C | |
| Beyond AIME | 69.1% | 03 | 5 | 50.0% | C | |
| MMLU | 90.6% | 06 | 100 | 95.0% | C | |
| Arena-Hard v2 | 71.0% | 07 | 16 | 60.0% | C | |
| HMMT25 | 85.8% | 10 | 25 | 62.5% | C | |
| AIME 2025 | 96.7% | 17 | 114 | 85.8% | C | |
| HMMT 2025 | 85.8% | 21 | 33 | 37.5% | C | |
| LiveCodeBench v6 | 71.7% | 27 | 53 | 50.0% | C | |
| IFEval | 84.8% | 39 | 65 | 40.6% | C | |
| MMLU-Pro | 81.7% | 41 | 129 | 68.8% | C | |
| BrowseComp | 49.5% | 45 | 58 | 22.8% | C | |
| Humanity's Last Exam | 11.2% | 79 | 93 | 15.2% | C | |
| SWE-Bench Verified | 45.0% | 91 | 105 | 13.5% | C | |
| GPQA | 78.7% | 94 | 234 | 60.1% | C |
Sarvam 105B Arena Results
Preference and agent-evaluation results for the default version.
No Arena results
The default version does not have a matching Arena result yet.
Sarvam 105B Runtime Performance
Provider-specific output speed and catalog latency for Sarvam-105B. Runtime does not affect the capability score.
No runtime data
Browse runtime rankingsNo provider-specific speed or latency record is linked to the default version yet.
Output Speed is generated output tokens received per second. Catalog latency is reported separately from observed provider TTFT.
Sarvam 105B Pricing
Official vendor API pricing appears first, followed by individual provider offers.
- Official API
- N/A
- Official provider
- N/A
- Lowest third-party
- From $0.04 input, $0.16 output per 1M via FastRouter
- Tracked offerings
- 3
| FastRouter | sarvam/sarvam-105b | global | $0.04 | $0.16 | 131.1K | |
| NanoGPT | sarvam-105b | global | $0.045 | $0.177 | 131.1K |
Official prices use only the vendor's configured official Provider and positive standard USD PAYG rates. Third-party offers remain explicitly labeled.
Sarvam 105B Versions
Available versions of this model. The score column identifies the version used in the overall ranking.
| Sarvam-105B | 44.6 | 105B | 131.1K | 131.1K | Yes | Apache 2.0 |
What is Sarvam 105B?
Key information about Sarvam 105B and its available data.
Sarvam-105B is Sarvam AI's flagship open-source Mixture-of-Experts reasoning model built for complex reasoning, coding, and agentic workflows.
It uses 128 sparse experts with Multi-head Latent Attention for efficient long-context inference and was pre-trained on 12 trillion tokens spanning code, mathematics, multilingual, and web data.
Data as of 2026-08-11.
Sarvam 105B vs nearby models
Open a comparison with the three ranked models immediately above and below this model.
Models similar to Sarvam 105B
Recommendations prioritize the same model type and family, then the closest LLMBoard score.
SA
Sarvam 30B
Sarvam AI
NV
Nemotron 3 Super
NVIDIA
LA
K EXAONE 236B A23B
LG AI Research
OP
GPT-5-mini
OpenAI
DE
DeepSeek-R1
DeepSeek
OP
o4 mini
OpenAI
FAQ
Common questions about Sarvam 105B.
When was Sarvam 105B released?
Sarvam 105B's default version was released on Mar 6, 2026.
How much does Sarvam 105B cost?
No official standard PAYG price is currently available for Sarvam 105B. The lowest tracked third-party offer starts at $0.04 input and $0.16 output via FastRouter.
Who created Sarvam 105B?
Sarvam 105B was created by Sarvam AI.
What is the context window for Sarvam 105B?
The default version has a 131.1K token context window.
Is Sarvam 105B open weight?
Yes. The default version is marked as open weight under Apache 2.0.
How many API providers offer Sarvam 105B?
3 provider offerings are linked to the default version.
What models should I compare Sarvam 105B with?
Nearby ranked alternatives include GPT-5-mini, K EXAONE 236B A23B, Nemotron 3 Super.