DeepSeek-V3 vs Sarvam 105B
Wins 3 of 6 areas
Facts · Long documents · Following instructions
Wins 2 of 6 areas
Agents · Reasoning
DeepSeek-V3 wins more areas, narrowly.Sarvam 105B is cheaper and better at reasoning.
Scores updated · 21 tests both models report · How we compare
Where each one wins
Tests won in each of the six areas where both have results. Each piece is one test, so a longer bar means more evidence; grey means the two scored within a point of each other.
- FactsGetting facts right instead of making them up20DeepSeek-V32 of 2 tests
- Long documentsFinding answers in very long texts10DeepSeek-V31 of 1 test
- Following instructionsDoing exactly what it is asked10DeepSeek-V31 of 1 test
- ReasoningHard problems that need careful thinking02Sarvam 105B2 of 3 tests · 1 tie
- AgentsCarrying out multi-step tasks on its own01Sarvam 105B1 of 1 test
- CodingWriting and fixing software22Even2 each
Agents, long documents and following instructions rest on a single test each.
What it costs
Prices per million tokens, roughly 750,000 words. The bars show the cost of a million tokens read plus a million written.
Sarvam 105B costs 81% less for the same work.
The biggest differences
The three tests each model wins by the widest margin. Scores are out of 100.
Where DeepSeek-V3 pulls ahead
- Hard command-line tasks in a real terminalTerminal-Bench Hard+13.7points ahead
- Code for real scientific research problemsSciCode+12.6points ahead
- Answers hard knowledge questions correctlyAA-Omniscience · Accuracy+7.9points ahead
Where Sarvam 105B pulls ahead
- Recent programming contest problemsLiveCodeBench v6+24.8points ahead
- Graduate-level biology, physics and chemistry questionsGPQA Diamond+8.3points ahead
- Very hard expert questions across many subjectsHumanity's Last Exam+6.3points ahead
Every test, side by side
All 21 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingEven
- LiveCodeBench v6Sarvam 105B by 24.846.971.7+24.8
- Terminal-Bench HardDeepSeek-V3 by 13.715.21.5+13.7
- SciCodeDeepSeek-V3 by 12.63926.4+12.6
- SWE-bench VerifiedSarvam 105B by 34245+3
ReasoningSarvam 105B
- GPQA DiamondSarvam 105B by 8.365.573.8+8.3
- Humanity's Last ExamSarvam 105B by 6.34.711+6.3
- CritPttie00tie
FactsDeepSeek-V3
- AA-Omniscience · AccuracyDeepSeek-V3 by 7.925.417.6+7.9
- AA-Omniscience · Non-hallucinationDeepSeek-V3 by 7.514.16.6+7.5
Following instructionsDeepSeek-V3
- IFBenchDeepSeek-V3 by 6.64134.4+6.6
Other results9 tests, not counted
Tests outside the eight areas. They are not counted above: several are summary scores built from other tests, or the same test under another name.
- HMMT 2025Sarvam 105B by 58.327.585.8+58.3
- AIME 2025Sarvam 105B by 45.451.396.7+45.4
- AA-OmniscienceDeepSeek-V3 by 18.7-40.7-59.4+18.7
- Artificial Analysis Coding IndexDeepSeek-V3 by 13.2239.8+13.2
- MATH-500 (EM)Sarvam 105B by 8.490.298.6+8.4
- MMLU-ProSarvam 105B by 5.875.981.7+5.8
- IFEvalDeepSeek-V3 by 1.386.184.8+1.3
- AA Intelligencetie9.78.8tie
- τ²-Bench Telecom (AA run)tie47.146.8tie
Questions people ask
Which is better, DeepSeek-V3 or Sarvam 105B?
DeepSeek-V3 wins three of the six areas where both have results: facts, long documents and following instructions. Sarvam 105B wins agents and reasoning, and costs 81% less. They are level on coding.
Which is better for coding?
Neither. They win 2 coding tests each of the 4 both models report.
Which is cheaper?
DeepSeek-V3 costs $0.24 per million input tokens and $0.90 per million output tokens; Sarvam 105B costs $0.04 and $0.17. That makes Sarvam 105B about 81% cheaper for the same work.
How do you compare the two?
We use the 21 benchmark tests both models have published scores on. The verdict counts the 12 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 9 are listed but not counted, because several are summary scores or repeat a test. Each score is the one shown on the model's own page.