Claude Opus 5.5 vs GPT-5.5
Wins 6 of 8 areas
Coding · Agents · Reasoning · Facts · Images and charts · Math
Wins 1 of 8 areas
Following instructions
Claude Opus 5.5 is the stronger all-rounder.GPT-5.5 is better at following instructions.
Both rank among the ten best models we track in reasoning, coding and math.
Scores updated · 47 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.
- CodingWriting and fixing software50Claude Opus 5.55 of 5 tests
- ReasoningHard problems that need careful thinking50Claude Opus 5.55 of 5 tests
- FactsGetting facts right instead of making them up30Claude Opus 5.53 of 4 tests · 1 tie
- MathCompetition and research-level math30Claude Opus 5.53 of 3 tests
- AgentsCarrying out multi-step tasks on its own31Claude Opus 5.53 of 4 tests
- Images and chartsUnderstanding pictures, charts and video20Claude Opus 5.52 of 2 tests
- Following instructionsDoing exactly what it is asked01GPT-5.51 of 1 test
- Long documentsFinding answers in very long texts00Even0 each · 1 tie
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.
Claude Opus 5.5 costs 31% 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 Claude Opus 5.5 pulls ahead
- Complex command-line tasks across many fieldsTerminal-Bench 4.0+45points ahead
- Long, multi-file coding tasks in real codebasesSWE-bench Pro+31.3points ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+30.4points ahead
Where GPT-5.5 pulls ahead
- Finds the root cause of IT system incidentsAA IT-Bench SRE+7.6points ahead
- Follows detailed instructions exactlyLiveBench · Instruction Following+5points ahead
Every test, side by side
All 47 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingClaude Opus 5.5
- LMArena · WebDevClaude Opus 5.5 by 328 rating points18151487+328 rating
- SWE-bench ProClaude Opus 5.5 by 31.389.958.6+31.3
- LiveBench · Agentic CodingClaude Opus 5.5 by 17.771.754+17.7
- SciCodeClaude Opus 5.5 by 11.166.955.8+11.1
- LiveBench · CodingClaude Opus 5.5 by 7.189.382.2+7.1
AgentsClaude Opus 5.5
- Terminal-Bench 4.0Claude Opus 5.5 by 4559.614.6+45
- GDPValClaude Opus 5.5 by 25.868.342.5+25.8
- AA IT-Bench SREGPT-5.5 by 7.638.245.8+7.6
- BrowseCompClaude Opus 5.5 by 6.490.884.4+6.4
ReasoningClaude Opus 5.5
- SimpleBenchClaude Opus 5.5 by 19.488.469+19.4
- Humanity's Last ExamClaude Opus 5.5 by 15.661.445.8+15.6
- ARC-AGI-2Claude Opus 5.5 by 6.791.785+6.7
- CritPtClaude Opus 5.5 by 4.631.727.1+4.6
- LiveBench · ReasoningClaude Opus 5.5 by 2.592.289.7+2.5
FactsClaude Opus 5.5
- AA-Omniscience · Non-hallucinationClaude Opus 5.5 by 30.441.411+30.4
- SimpleQA VerifiedClaude Opus 5.5 by 9.272.263+9.2
- AA-Omniscience · AccuracyClaude Opus 5.5 by 8.266.258+8.2
- Vectara HHEM hallucination ratelower is bettertie9.99.3tie
Images and chartsClaude Opus 5.5
- MMMU-ProClaude Opus 5.5 by 7.887.779.9+7.8
- LMArena · VisionClaude Opus 5.5 by 29 rating points13241295+29 rating
MathClaude Opus 5.5
- FrontierMath Tier 4Claude Opus 5.5 by 22.59572.5+22.5
- FrontierMath Tiers 1-3 (v2)Claude Opus 5.5 by 5.991.285.3+5.9
- LiveBench · MathematicsClaude Opus 5.5 by 1.297.195.9+1.2
Following instructionsGPT-5.5
- LiveBench · Instruction FollowingGPT-5.5 by 565.770.7+5
Other results22 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-OmniscienceClaude Opus 5.5 by 25.946.420.5+25.9
- PostTrainBenchClaude Opus 5.5 by 20.949.328.4+20.9
- Program BenchClaude Opus 5.5 by 20.491.270.8+20.4
- HiL-Bench (Tools-allowed)Claude Opus 5.5 by 19.35939.7+19.3
- AA IntelligenceClaude Opus 5.5 by 19.257.638.4+19.2
- HLE (with tools)Claude Opus 5.5 by 15.567.752.2+15.5
- OfficeQA ProClaude Opus 5.5 by 13.667.754.1+13.6
- HealthBench ProfessionalClaude Opus 5.5 by 8.159.951.8+8.1
- HealthBench Professional length-adjustedClaude Opus 5.5 by 8.159.951.8+8.1
- DeepSWE 1.1Claude Opus 5.5 by 7.274.267+7.2
- vectara_avg_summary_lengthClaude Opus 5.5 by 62.5 rating points192.1129.6+62.5 rating
- Legal Agent BenchmarkClaude Opus 5.5 by 6.28.32.1+6.2
- GDPval-AA (Elo)Claude Opus 5.5 by 51 rating points18201769+51 rating
- Toolathlon VerifiedClaude Opus 5.5 by 4.377.873.5+4.3
- ARC-AGI-1Claude Opus 5.5 by 2.597.595+2.5
- GDP (Surge AI)Claude Opus 5.5 by 1.326.224.9+1.3
- livebench_data_analysisGPT-5.5 by 1.380.381.6+1.3
- livebench_languageGPT-5.5 by 1.186.387.4+1.1
- vectara_factual_consistencytie90.190.7tie
- HealthBenchtie5756.5tie
- HealthBench length-adjustedtie5756.5tie
- vectara_answer_ratetie99.7100tie
Questions people ask
Which is better, Claude Opus 5.5 or GPT-5.5?
Claude Opus 5.5 wins six of the eight areas we test: coding, agents, reasoning, facts, images and charts and math. GPT-5.5 wins following instructions. They are level on long documents.
Which is better for coding?
Claude Opus 5.5. It wins 5 of the 5 coding tests both models report; GPT-5.5 wins none.
Which is cheaper?
Claude Opus 5.5 costs $4.00 per million input tokens and $20.00 per million output tokens; GPT-5.5 costs $5.00 and $30.00. That makes Claude Opus 5.5 about 31% cheaper for the same work.
How do you compare the two?
We use the 47 benchmark tests both models have published scores on. The verdict counts the 25 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 22 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.