Benchmarks of Sarvam 30B10
BenchmarkByScoreRank
Overall1
Knowledge4
Science2
Coding1
Legacy · saturated or no longer reported2
No benchmark matches
—per task?
price not verifiedprice
—output speed?
66Kcontext
Ranking?
Alternatives?
ModelScoreCostPer taskWhat changes
BetterGPT-6 Lunalow×3.3—$0.005Function calling only with none · Chat Completions+3
OpenQwen3.8 27Bxhigh×5.1—$1.01max_tokens range · QwenCloud+4
BetterGranite 4.2 3BthinkingOpen+38%—$0.006
BetterGemma 4 E4BthinkingOpen+36%—$0.008BetterGPT-5.6 Lunalow×3.2—$0.01ModelScoreCostPer taskWhat changes
Nothing here yet.
ModelScoreCostPer taskWhat changes
GPT-6 Lunalow×3.3—$0.005Function calling only with none · Chat Completions+3Granite 4.2 3BthinkingOpen+38%—$0.006Gemma 4 E4BthinkingOpen+36%—$0.008GPT-5.6 Lunalow×3.2—$0.01GPT OSS 20BhighOpen+37%—$0.011ModelScoreCostPer taskWhat changes
Qwen3.8 27BxhighOpen×5.1—$1.01max_tokens range · QwenCloud+4IK2 Horizon MoVA 36B A4BthinkingOpen×3.9—$1 / $4Memory (4-bit): 19 GB → 22 GBQwen 3.5 27BthinkingOpen×3.5—$0.30 / $2.40AG9v3 39A5BOpen×3.3——Memory (4-bit): 19 GB → 23 GBQwen3.6 27BthinkingOpen×3.3—$0.62ModelScoreCostPer taskWhat changes
SSarvam 105BOpen+34%—$15 / $60Memory (4-bit): 19 GB → 64 GBSSarvam MOpen-19%—$0.70 / $1.16Context: 66K → 8KAll Sarvam models →Run it yourself
estimateFits on✗ 8 GB (phone, laptop)✗ 16 GB✓ 24 GB GPU · compressed✓ 48 GB GPU · 8-bit✓ 80 GB GPU · full quality
LicenceApache-2.0free for commercial use
Benchmarks
all categoriesChat & Knowledge
Humanity's Last Exam7.5#216
AA Omniscience Index-71.5#296
CritPt (physics)0.3#149
AA Intelligence Index 4.3.2≈6.6#286
AA Omniscience · Accuracy12.6#262
GPQA Diamond63.3#217
Coding
Agents & Tool Use
τ²-bench · AA34.5#137
Writing & Language
IFBench (instructions)26.5#225
FAQ
How much does Sarvam 30B cost?›
There is no per-unit API price: price not verified.
How good is Sarvam 30B, and what is it best at?›
#214 of 334 in Chat & Knowledge, #181 of 239 in Coding, #135 of 242 in Agents & Tool Use, #222 of 245 in Writing & Language.
Is Sarvam 30B better than Sarvam M?›
On AA Intelligence Index 4.3.2, Sarvam 30B scores 6.6 vs 5.3 — higher.
What is the context window of Sarvam 30B?›
66K tokens.
When was Sarvam 30B released? Is there a newer version?›
Released 2026-03-06. It is the current version in its line.
Is Sarvam 30B open source?›
Yes, the weights are open (Apache-2.0 licence).
Can I run Sarvam 30B locally?›
The weights are open and can be self-hosted; hardware needs are not confirmed yet.
Can Sarvam 30B read images?›
No, it accepts text only.
What is Sarvam 30B?›
Sarvam 30B is an open-weights language model from Sarvam, released Mar 2026, the successor to Sarvam M.
What are the best alternatives to Sarvam 30B?›
GPT-6 Luna (better), Qwen3.8 27B (open weights).