DeepSeek-V3.1 vs Gemini 2.5 Pro
Wins 2 of 6 areas
Agents · Facts
Wins 4 of 6 areas
Coding · Reasoning · Long documents · Following instructions
Gemini 2.5 Pro is the stronger all-rounder.DeepSeek-V3.1 is cheaper and better at agents.
Scores updated · 33 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.
- ReasoningHard problems that need careful thinking03Gemini 2.5 Pro3 of 4 tests · 1 tie
- Following instructionsDoing exactly what it is asked02Gemini 2.5 Pro2 of 2 tests
- CodingWriting and fixing software12Gemini 2.5 Pro2 of 3 tests
- Long documentsFinding answers in very long texts01Gemini 2.5 Pro1 of 1 test
- AgentsCarrying out multi-step tasks on its own20DeepSeek-V3.12 of 2 tests
- FactsGetting facts right instead of making them up21DeepSeek-V3.12 of 3 tests
Long documents rests on a single test.
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.
DeepSeek-V3.1 costs 80% less for the same work.
The biggest differences
The three tests each model wins by the widest margin. Scores are out of 100.
Where Gemini 2.5 Pro pulls ahead
- Reasons across sets of long documentsAA-LCR+12.3points ahead
- Common-sense trick questionsSimpleBench+11.6points ahead
- Answers hard knowledge questions correctlyAA-Omniscience · Accuracy+10.1points ahead
Where DeepSeek-V3.1 pulls ahead
- Finds hard-to-locate facts by browsing the webBrowseComp+20.1points ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+8.4points ahead
- Fixes real GitHub issues in Python projectsSWE-bench Verified+2.8points ahead
Every test, side by side
All 33 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingGemini 2.5 Pro
- SciCodeGemini 2.5 Pro by 7.239.146.3+7.2
- SWE-bench VerifiedDeepSeek-V3.1 by 2.86663.2+2.8
- Terminal-Bench HardGemini 2.5 Pro by 1.52526.5+1.5
AgentsDeepSeek-V3.1
- BrowseCompDeepSeek-V3.1 by 20.1309.9+20.1
- GDPValDeepSeek-V3.1 by 5.65.60+5.6
ReasoningGemini 2.5 Pro
- SimpleBenchGemini 2.5 Pro by 11.64051.6+11.6
- GPQA DiamondGemini 2.5 Pro by 5.777.983.6+5.7
- Humanity's Last ExamGemini 2.5 Pro by 3.714.318+3.7
- CritPttie22.6tie
FactsDeepSeek-V3.1
- AA-Omniscience · AccuracyGemini 2.5 Pro by 10.12939+10.1
- AA-Omniscience · Non-hallucinationDeepSeek-V3.1 by 8.417.59.1+8.4
- Vectara HHEM hallucination ratelower is betterDeepSeek-V3.1 by 1.55.57+1.5
Following instructionsGemini 2.5 Pro
- Multi-ChallengeGemini 2.5 Pro by 7.546.153.6+7.5
- IFBenchGemini 2.5 Pro by 7.241.548.7+7.2
Other results18 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.
- SimpleQADeepSeek-V3.1 by 42.693.450.8+42.6
- AIME 2025Gemini 2.5 Pro by 33.249.883+33.2
- HMMT 2025Gemini 2.5 Pro by 32.233.565.7+32.2
- LiveCodeBenchGemini 2.5 Pro by 17.856.474.2+17.8
- Artificial Analysis Coding IndexGemini 2.5 Pro by 1729.746.7+17
- BrowseComp-ZHDeepSeek-V3.1 by 1749.232.2+17
- τ²-Bench Telecom (AA run)Gemini 2.5 Pro by 16.737.454.1+16.7
- AA Agentic IndexDeepSeek-V3.1 by 15.418.93.5+15.4
- AA-OmniscienceGemini 2.5 Pro by 13.3-29.6-16.4+13.3
- Aider-PolyglotGemini 2.5 Pro by 8.168.476.5+8.1
- Terminal-BenchDeepSeek-V3.1 by 631.325.3+6
- HMMT Feb. 2025DeepSeek-V3.1 by 585.880.8+5
- vectara_answer_rateGemini 2.5 Pro by 4.694.599.1+4.6
- vectara_avg_summary_lengthGemini 2.5 Pro by 42.7 rating points63.7106.4+42.7 rating
- MMLU-ProGemini 2.5 Pro by 2.383.786+2.3
- AA IntelligenceGemini 2.5 Pro by 1.513.515+1.5
- vectara_factual_consistencyDeepSeek-V3.1 by 1.594.593+1.5
- AIME 2024DeepSeek-V3.1 by 1.193.192+1.1
Questions people ask
Which is better, DeepSeek-V3.1 or Gemini 2.5 Pro?
Gemini 2.5 Pro wins four of the six areas where both have results: coding, reasoning, long documents and following instructions. DeepSeek-V3.1 wins agents and facts, and costs 80% less.
Which is better for coding?
Gemini 2.5 Pro. It wins 2 of the 3 coding tests both models report; DeepSeek-V3.1 wins 1.
Which is cheaper?
DeepSeek-V3.1 costs $0.56 per million input tokens and $1.68 per million output tokens; Gemini 2.5 Pro costs $1.25 and $10.00. That makes DeepSeek-V3.1 about 80% cheaper for the same work.
How do you compare the two?
We use the 33 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 18 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.