DeepSeek-V3 vs Gemma 3 27B
Wins 6 of 6 areas
Coding · Agents · Reasoning · Facts · Long documents · Following instructions
Wins 0 of 6 areas
—
DeepSeek-V3 is the stronger all-rounder.Gemma 3 27B is cheaper.
Scores updated · 36 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.
- CodingWriting and fixing software40DeepSeek-V34 of 4 tests
- FactsGetting facts right instead of making them up30DeepSeek-V33 of 3 tests
- AgentsCarrying out multi-step tasks on its own10DeepSeek-V31 of 3 tests · 2 ties
- ReasoningHard problems that need careful thinking10DeepSeek-V31 of 3 tests · 2 ties
- Long documentsFinding answers in very long texts10DeepSeek-V31 of 1 test
- Following instructionsDoing exactly what it is asked10DeepSeek-V31 of 1 test
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.
Gemma 3 27B costs 79% less for the same work.
The biggest differences
The tests each model wins by the widest margin, up to three each. Scores are out of 100.
Where DeepSeek-V3 pulls ahead
- Reasons across sets of long documentsAA-LCR+33.4points ahead
- Graduate-level biology, physics and chemistry questionsGPQA Diamond+22.7points ahead
- Recent programming contest problemsLiveCodeBench v6+17.8points ahead
Where Gemma 3 27B pulls ahead
No clear win on a test scored out of 100.
Every test, side by side
All 36 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingDeepSeek-V3
- LiveCodeBench v6DeepSeek-V3 by 17.846.929.1+17.8
- SciCodeDeepSeek-V3 by 15.73923.3+15.7
- Terminal-Bench 2.1DeepSeek-V3 by 12.416.94.5+12.4
- Terminal-Bench HardDeepSeek-V3 by 11.415.23.8+11.4
AgentsDeepSeek-V3
- τ-Bench V3 · BankingDeepSeek-V3 by 3.94.70.8+3.9
- GDPValtie00tie
- Terminal-Bench 4.0tie00tie
ReasoningDeepSeek-V3
- GPQA DiamondDeepSeek-V3 by 22.765.542.8+22.7
- Humanity's Last Examtie4.74.4tie
- CritPttie00tie
FactsDeepSeek-V3
- AA-Omniscience · AccuracyDeepSeek-V3 by 12.525.412.9+12.5
- AA-Omniscience · Non-hallucinationDeepSeek-V3 by 6.214.17.9+6.2
- Vectara HHEM hallucination ratelower is betterDeepSeek-V3 by 1.36.17.4+1.3
Following instructionsDeepSeek-V3
- IFBenchDeepSeek-V3 by 9.24131.8+9.2
Other results21 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.
- Aider-PolyglotDeepSeek-V3 by 44.749.64.9+44.7
- τ²-Bench Telecom (AA run)DeepSeek-V3 by 36.647.110.5+36.6
- AA-OmniscienceDeepSeek-V3 by 26.5-40.7-67.2+26.5
- HumanEvalGemma 3 27B by 22.665.287.8+22.6
- LiveCodeBenchDeepSeek-V3 by 19.549.229.7+19.5
- τ²-BenchDeepSeek-V3 by 16.332.516.2+16.3
- SimpleQADeepSeek-V3 by 14.924.910+14.9
- vectara_avg_summary_lengthGemma 3 27B by 14.781.796.4+14.7
- MMLU-ProXDeepSeek-V3 by 13.970.556.6+13.9
- Artificial Analysis Coding IndexDeepSeek-V3 by 12.92310.1+12.9
- MMLUDeepSeek-V3 by 11.688.576.9+11.6
- MMMLUDeepSeek-V3 by 8.779.470.7+8.7
- MMLU-ProDeepSeek-V3 by 8.475.967.5+8.4
- AA IntelligenceDeepSeek-V3 by 4.89.74.9+4.8
- IFEvalGemma 3 27B by 4.386.190.4+4.3
- GSM8KGemma 3 27B by 1.99495.9+1.9
- vectara_answer_rateGemma 3 27B by 1.397.598.8+1.3
- vectara_factual_consistencyDeepSeek-V3 by 1.393.992.6+1.3
- MATHDeepSeek-V3 by 1.290.289+1.2
- MBPPDeepSeek-V3 by 175.474.4+1
- BBHtie87.587.6tie
Questions people ask
Which is better, DeepSeek-V3 or Gemma 3 27B?
DeepSeek-V3 wins all six areas where both have results: coding, agents, reasoning, facts, long documents and following instructions. Gemma 3 27B wins none, but costs 79% less.
Which is better for coding?
DeepSeek-V3. It wins 4 of the 4 coding tests both models report; Gemma 3 27B wins none.
Which is cheaper?
DeepSeek-V3 costs $0.24 per million input tokens and $0.90 per million output tokens; Gemma 3 27B costs $0.08 and $0.16. That makes Gemma 3 27B about 79% cheaper for the same work.
How do you compare the two?
We use the 36 benchmark tests both models have published scores on. The verdict counts the 15 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 21 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.