DeepSeek-V3.2 vs Gemini 3.1 Pro
Wins 0 of 7 areas
—
Wins 7 of 7 areas
Coding · Agents · Reasoning · Facts · Math · Long documents · Following instructions
Gemini 3.1 Pro is the stronger all-rounder.DeepSeek-V3.2 is cheaper.
Scores updated · 49 tests both models report · How we compare
Where each one wins
Tests won in each of the seven 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 software08Gemini 3.1 Pro8 of 8 tests
- AgentsCarrying out multi-step tasks on its own05Gemini 3.1 Pro5 of 5 tests
- ReasoningHard problems that need careful thinking04Gemini 3.1 Pro4 of 4 tests
- MathCompetition and research-level math03Gemini 3.1 Pro3 of 3 tests
- FactsGetting facts right instead of making them up13Gemini 3.1 Pro3 of 4 tests
- Long documentsFinding answers in very long texts01Gemini 3.1 Pro1 of 1 test
- Following instructionsDoing exactly what it is asked01Gemini 3.1 Pro1 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.
DeepSeek-V3.2 costs 95% 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 Gemini 3.1 Pro pulls ahead
- Abstract visual puzzles that people can solveARC-AGI-2+73.1points ahead
- Short factual questions, answered correctlySimpleQA Verified+46points ahead
- Long, multi-file coding tasks in real codebasesSWE-bench Pro+38.6points ahead
Where DeepSeek-V3.2 pulls ahead
No clear win on a test scored out of 100.
Every test, side by side
All 49 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingGemini 3.1 Pro
- SWE-bench ProGemini 3.1 Pro by 38.615.654.2+38.6
- Terminal-Bench 2.1Gemini 3.1 Pro by 2746.873.8+27
- SciCodeGemini 3.1 Pro by 19.838.958.7+19.8
- Terminal-Bench HardGemini 3.1 Pro by 18.235.653.8+18.2
- LiveCodeBench v6Gemini 3.1 Pro by 8.483.391.7+8.4
- LMArena · WebDevGemini 3.1 Pro by 84 rating points13621446+84 rating
- SWE-bench VerifiedGemini 3.1 Pro by 7.573.180.6+7.5
- SWE-bench MultilingualGemini 3.1 Pro by 6.770.276.9+6.7
AgentsGemini 3.1 Pro
- BrowseCompGemini 3.1 Pro by 34.551.485.9+34.5
- AA ApexAgentsGemini 3.1 Pro by 17.514.532+17.5
- MCP AtlasGemini 3.1 Pro by 762.269.2+7
- GDPValGemini 3.1 Pro by 4.99.814.7+4.9
- τ-Bench V3 · BankingGemini 3.1 Pro by 2.618.821.4+2.6
ReasoningGemini 3.1 Pro
- ARC-AGI-2Gemini 3.1 Pro by 73.1477.1+73.1
- Humanity's Last ExamGemini 3.1 Pro by 22.424.647+22.4
- CritPtGemini 3.1 Pro by 14.82.917.7+14.8
- GPQA DiamondGemini 3.1 Pro by 10.18494.1+10.1
FactsGemini 3.1 Pro
- SimpleQA VerifiedGemini 3.1 Pro by 4627.573.5+46
- AA-Omniscience · Non-hallucinationGemini 3.1 Pro by 31.817.349.1+31.8
- AA-Omniscience · AccuracyGemini 3.1 Pro by 21.93354.9+21.9
- Vectara HHEM hallucination ratelower is betterDeepSeek-V3.2 by 4.16.310.4+4.1
MathGemini 3.1 Pro
- HMMT Feb. 2026Gemini 3.1 Pro by 10.684.194.7+10.6
- AIME 2026Gemini 3.1 Pro by 4.194.298.3+4.1
- IMOAnswerBenchGemini 3.1 Pro by 2.778.381+2.7
Following instructionsGemini 3.1 Pro
- IFBenchGemini 3.1 Pro by 16.460.777.1+16.4
Other results23 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.
- AA-OmniscienceGemini 3.1 Pro by 54.4-22.531.9+54.4
- ARC-AGI-1Gemini 3.1 Pro by 415798+41
- Artificial Analysis Coding IndexGemini 3.1 Pro by 24.644.268.8+24.6
- Terminal-Bench 2.0Gemini 3.1 Pro by 22.146.468.5+22.1
- CyberGymGemini 3.1 Pro by 21.517.338.8+21.5
- DeepSearchQA (F1)Gemini 3.1 Pro by 2160.981.9+21
- t2-benchGemini 3.1 Pro by 19.180.299.3+19.1
- browsecomp_with_context_managerGemini 3.1 Pro by 18.367.685.9+18.3
- MCPMarkGemini 3.1 Pro by 17.93855.9+17.9
- frontiermath_tier_4_v1Gemini 3.1 Pro by 14.62.116.7+14.6
- Tool-DecathlonGemini 3.1 Pro by 13.635.248.8+13.6
- ToolathlonGemini 3.1 Pro by 13.635.248.8+13.6
- HLE (with tools)Gemini 3.1 Pro by 10.640.851.4+10.6
- LiveCodeBenchGemini 3.1 Pro by 8.483.391.7+8.4
- AA IntelligenceGemini 3.1 Pro by 8.221.529.7+8.2
- AA Agentic IndexDeepSeek-V3.2 by 818.310.3+8
- vectara_answer_rateGemini 3.1 Pro by 6.892.699.4+6.8
- MMLU-ProGemini 3.1 Pro by 68591+6
- τ²-Bench Telecom (AA run)Gemini 3.1 Pro by 590.695.6+5
- HMMT Nov. 2025Gemini 3.1 Pro by 4.89094.8+4.8
- vectara_avg_summary_lengthGemini 3.1 Pro by 45.7 rating points62107.7+45.7 rating
- vectara_factual_consistencyDeepSeek-V3.2 by 4.193.789.6+4.1
- τ³-BenchDeepSeek-V3.2 by 2.169.267.1+2.1
Questions people ask
Which is better, DeepSeek-V3.2 or Gemini 3.1 Pro?
Gemini 3.1 Pro wins all seven areas where both have results: coding, agents, reasoning, facts, math, long documents and following instructions. DeepSeek-V3.2 wins none, but costs 95% less.
Which is better for coding?
Gemini 3.1 Pro. It wins 8 of the 8 coding tests both models report; DeepSeek-V3.2 wins none.
Which is cheaper?
DeepSeek-V3.2 costs $0.28 per million input tokens and $0.42 per million output tokens; Gemini 3.1 Pro costs $2.00 and $12.00. That makes DeepSeek-V3.2 about 95% cheaper for the same work.
How do you compare the two?
We use the 49 benchmark tests both models have published scores on. The verdict counts the 26 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 23 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.