Claude Opus 4.6 vs Gemini 2.5 Pro
Wins 5 of 8 areas
Coding · Agents · Reasoning · Math · Long documents
Wins 0 of 8 areas
—
Claude Opus 4.6 is the stronger all-rounder.Gemini 2.5 Pro is cheaper.
Scores updated · 42 tests both models report · How we compare
Where each one wins
Tests won in each of the eight areas we test. 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 thinking50Claude Opus 4.65 of 5 tests
- CodingWriting and fixing software40Claude Opus 4.64 of 4 tests
- AgentsCarrying out multi-step tasks on its own20Claude Opus 4.62 of 2 tests
- MathCompetition and research-level math20Claude Opus 4.62 of 2 tests
- Long documentsFinding answers in very long texts10Claude Opus 4.61 of 1 test
- FactsGetting facts right instead of making them up22Even2 each
- Images and chartsUnderstanding pictures, charts and video11Even1 each · 1 tie
- Following instructionsDoing exactly what it is asked11Even1 each
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.
Gemini 2.5 Pro costs 63% less for the same work.
The biggest differences
The three tests each model wins by the widest margin. Scores are out of 100.
Where Claude Opus 4.6 pulls ahead
- Finds hard-to-locate facts by browsing the webBrowseComp+74.1points ahead
- Abstract visual puzzles that people can solveARC-AGI-2+63.9points ahead
- Unpublished advanced math problemsFrontierMath Tiers 1-3 (v2)+41.4points ahead
Where Gemini 2.5 Pro pulls ahead
- Keeps track of context across a multi-turn chatMulti-Challenge+16.4points ahead
- Reads and reasons about text in imagesOCRBenchv2+10.9points ahead
- Short factual questions, answered correctlySimpleQA Verified+9points ahead
Every test, side by side
All 42 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingClaude Opus 4.6
- LMArena · WebDevClaude Opus 4.6 by 322 rating points15461224+322 rating
- Terminal-Bench HardClaude Opus 4.6 by 19.746.226.5+19.7
- SWE-bench VerifiedClaude Opus 4.6 by 17.680.863.2+17.6
- SciCodeClaude Opus 4.6 by 5.651.946.3+5.6
AgentsClaude Opus 4.6
- BrowseCompClaude Opus 4.6 by 74.1849.9+74.1
- GDPValClaude Opus 4.6 by 55.955.90+55.9
ReasoningClaude Opus 4.6
- ARC-AGI-2Claude Opus 4.6 by 63.968.84.9+63.9
- Humanity's Last ExamClaude Opus 4.6 by 21.939.918+21.9
- SimpleBenchClaude Opus 4.6 by 1667.651.6+16
- CritPtClaude Opus 4.6 by 1012.62.6+10
- GPQA DiamondClaude Opus 4.6 by 689.683.6+6
FactsEven
- AA-Omniscience · Non-hallucinationClaude Opus 4.6 by 28.137.29.1+28.1
- SimpleQA VerifiedGemini 2.5 Pro by 94756+9
- AA-Omniscience · AccuracyClaude Opus 4.6 by 7.94739+7.9
- Vectara HHEM hallucination ratelower is betterGemini 2.5 Pro by 5.212.27+5.2
Images and chartsEven
- OCRBenchv2Gemini 2.5 Pro by 10.948.459.3+10.9
- LMArena · VisionClaude Opus 4.6 by 53 rating points13161263+53 rating
- MMMU-Protie75.474.9tie
MathClaude Opus 4.6
- FrontierMath Tiers 1-3 (v2)Claude Opus 4.6 by 41.46624.6+41.4
- FrontierMath Tier 4Claude Opus 4.6 by 26.826.80+26.8
Following instructionsEven
- Multi-ChallengeGemini 2.5 Pro by 16.437.153.6+16.4
- IFBenchClaude Opus 4.6 by 4.453.148.7+4.4
Other results19 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 Agentic IndexClaude Opus 4.6 by 64.167.63.5+64.1
- ARC-AGI-1Claude Opus 4.6 by 569337+56
- τ²-Bench Telecom (AA run)Claude Opus 4.6 by 3892.154.1+38
- OpenAI-MRCR (1M)Claude Opus 4.6 by 34.192.958.8+34.1
- AA-OmniscienceClaude Opus 4.6 by 3013.7-16.4+30
- HLE (with tools)Claude Opus 4.6 by 24.953.328.4+24.9
- frontiermath_tier_4_v1Claude Opus 4.6 by 18.722.94.2+18.7
- AA IntelligenceClaude Opus 4.6 by 16.931.915+16.9
- AIME 2025Claude Opus 4.6 by 16.899.883+16.8
- HMMT Nov. 2025Claude Opus 4.6 by 16.396.380+16.3
- LiveCodeBenchClaude Opus 4.6 by 14.688.874.2+14.6
- MATH-VisionGemini 2.5 Pro by 7.865.573.3+7.8
- vectara_factual_consistencyGemini 2.5 Pro by 5.287.893+5.2
- matharena_visual_math_overallGemini 2.5 Pro by 4.972.377.2+4.9
- vectara_avg_summary_lengthClaude Opus 4.6 by 31.2 rating points137.6106.4+31.2 rating
- MMLU-ProClaude Opus 4.6 by 3.189.186+3.1
- OCRBench v2 (Chinese)Gemini 2.5 Pro by 2.459.862.2+2.4
- Artificial Analysis Coding IndexClaude Opus 4.6 by 1.448.146.7+1.4
- vectara_answer_ratetie99.899.1tie
Questions people ask
Which is better, Claude Opus 4.6 or Gemini 2.5 Pro?
Claude Opus 4.6 wins five of the eight areas we test: coding, agents, reasoning, math and long documents. Gemini 2.5 Pro wins none, but costs 63% less. They are level on facts, images and charts and following instructions.
Which is better for coding?
Claude Opus 4.6. It wins 4 of the 4 coding tests both models report; Gemini 2.5 Pro wins none.
Which is cheaper?
Claude Opus 4.6 costs $5.00 per million input tokens and $25.00 per million output tokens; Gemini 2.5 Pro costs $1.25 and $10.00. That makes Gemini 2.5 Pro about 63% cheaper for the same work.
How do you compare the two?
We use the 42 benchmark tests both models have published scores on. The verdict counts the 23 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 19 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.