Sarvam 105B
Sarvam 105B is behind the leaders in agentic tasks. Too few results yet to rate reasoning, coding, safety, long context, math, multimodal tasks, multilingual tasks, instruction following, or factuality.
Capability profile
Bars show the model's median result as a share of the leading model's, per capability. Select a row to see its results.
Too few results yet to rate Reasoning, Coding, Safety, Long Context, Math, Multimodal, Multilingual, Instruction Following or Factuality.
Price
$0.04input$0.17outputper million tokens
From Artificial Analysis · All prices
Evidence
27results on25benchmarks
- 15 aggregator
- 12 vendor-reported
From 2 sources · latest Oct 7, 2026 · How verification works
Research
1 paper reference Sarvam 105BSarvam 105B benchmark results
27 results on 25 benchmarks, grouped by capability. Each bar is the result as a share of the capability leader's; every result links to its source.
42.5% behind the leader2 of 7 ranked benchmarks measured
- 49.50Oct 6, 2026
- 11.89Jun 15, 2026
0 of 6 ranked benchmarks measured
- 0.00Oct 7, 2026
- GPQA Diamond73.84Oct 7, 2026gpqa
- Humanity's Last Exam11.03Oct 7, 2026aa_hle
Show 2 more reasoning resultsHide 2 reasoning results
- GPQA Diamond78.70Oct 6, 2026GPQA
- 11.20Oct 6, 2026
0 of 10 ranked benchmarks measured
- Terminal-Bench Hard1.52Oct 7, 2026aa_terminalbench_hard
- SciCode26.39Sep 4, 2026aa_scicode
- 45.00Oct 6, 2026
Show 1 more coding resultHide 1 coding result
- 71.70Oct 6, 2026
0 of 3 ranked benchmarks measured
- 0.00Oct 7, 2026
0 of 3 ranked benchmarks measured
- IFBench34.35Oct 7, 2026aa_ifbench
0 of 4 ranked benchmarks measured
- AA-Omniscience · Accuracy17.60Oct 7, 2026omniscienceAccuracy
- AA-Omniscience · Non-hallucination6.57Oct 7, 2026omniscienceNonHallucination
More results
Benchmarks outside the capability baskets. They are not ranked against a leader.
- τ²-Bench Telecom (AA run)46.78Oct 7, 2026aa_tau2
- AA Intelligence8.79Oct 7, 2026aa_intelligence_index
- -59.38Oct 7, 2026
- 9.81Jun 18, 2026
- 24.69Jun 18, 2026
- 69.10Oct 6, 2026
Show 6 more resultsHide 6 results
- 96.70Oct 6, 2026
- 71.00Oct 6, 2026
- HMMT 202585.80Oct 6, 2026HMMT25
- 84.80Aug 31, 2026
- MATH-500 (EM)98.60Oct 6, 2026MATH-500
- 81.70Oct 6, 2026
Sarvam 105B: common questions
Who makes Sarvam 105B?
Sarvam 105B is made by Sarvam AI.
When was Sarvam 105B released?
Sarvam 105B was released on Mar 6, 2026, according to Artificial Analysis.
What is Sarvam 105B good at?
Sarvam 105B is behind the leaders in agentic tasks. Too few results yet to rate reasoning, coding, safety, long context, math, multimodal tasks, multilingual tasks, instruction following, or factuality.
How much does Sarvam 105B cost?
Sarvam 105B costs $0.04 per million input tokens and $0.17 per million output tokens, according to Artificial Analysis. At a mix of three input tokens to one output token, it is cheaper than 94% of the 329 priced models we track.
How many benchmarks has Sarvam 105B been tested on?
We track 27 results for Sarvam 105B on 25 benchmarks from 2 sources. The latest was recorded on Oct 7, 2026.
About this record
Where Sarvam 105B's numbers come from, and every name it appears under.
- Tracked since
- May 2, 2026
- Newest source mention
- Jun 16, 2026
Where the results come from
Verification: 27 scores · 0 independently verified · 15 aggregator-attributed · 12 vendor-reported. How these tiers are assigned
From 2 sources on 2 sites. Artificial Analysis supplies 15 of them. Bars are coloured by trust tier.
- artificialanalysis.ai15
- api.llm-stats.com12