Claude Opus 5.5 vs GPT-5.4
Wins 7 of 8 areas
Coding · Agents · Reasoning · Facts · Images and charts · Math · Long documents
Wins 1 of 8 areas
Following instructions
Claude Opus 5.5 is the stronger all-rounder.GPT-5.4 is cheaper and better at following instructions.
Scores updated · 34 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 software60Claude Opus 5.56 of 6 tests
- ReasoningHard problems that need careful thinking40Claude Opus 5.54 of 4 tests
- FactsGetting facts right instead of making them up30Claude Opus 5.53 of 3 tests
- MathCompetition and research-level math30Claude Opus 5.53 of 3 tests
- AgentsCarrying out multi-step tasks on its own20Claude Opus 5.52 of 2 tests
- Images and chartsUnderstanding pictures, charts and video20Claude Opus 5.52 of 2 tests
- Long documentsFinding answers in very long texts10Claude Opus 5.51 of 1 test
- Following instructionsDoing exactly what it is asked01GPT-5.41 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.
GPT-5.4 costs 27% 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
- Extremely hard research-level math problemsFrontierMath Tier 4+46points ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+33.1points ahead
- Long, multi-file coding tasks in real codebasesSWE-bench Pro+32.2points ahead
Where GPT-5.4 pulls ahead
- Follows detailed instructions exactlyLiveBench · Instruction Following+4.5points ahead
Every test, side by side
All 34 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 343 rating points18071464+343 rating
- SWE-bench ProClaude Opus 5.5 by 32.289.957.7+32.2
- SWE-bench MultilingualClaude Opus 5.5 by 22.293.971.7+22.2
- LiveBench · Agentic CodingClaude Opus 5.5 by 17.971.753.8+17.9
- LiveBench · CodingClaude Opus 5.5 by 11.889.377.5+11.8
- SciCodeClaude Opus 5.5 by 10.366.956.6+10.3
AgentsClaude Opus 5.5
- GDPValClaude Opus 5.5 by 3168.337.3+31
- AA IT-Bench SREClaude Opus 5.5 by 3.738.234.5+3.7
ReasoningClaude Opus 5.5
- ARC-AGI-2Claude Opus 5.5 by 18.491.773.3+18.4
- Humanity's Last ExamClaude Opus 5.5 by 17.761.443.7+17.7
- CritPtClaude Opus 5.5 by 8.331.723.4+8.3
- LiveBench · ReasoningClaude Opus 5.5 by 4.192.288.1+4.1
FactsClaude Opus 5.5
- AA-Omniscience · Non-hallucinationClaude Opus 5.5 by 33.141.48.3+33.1
- SimpleQA VerifiedClaude Opus 5.5 by 27.172.245.1+27.1
- AA-Omniscience · AccuracyClaude Opus 5.5 by 15.366.250.9+15.3
Images and chartsClaude Opus 5.5
- MMMU-ProClaude Opus 5.5 by 9.387.778.4+9.3
- LMArena · VisionClaude Opus 5.5 by 18 rating points13211303+18 rating
MathClaude Opus 5.5
- FrontierMath Tier 4Claude Opus 5.5 by 469549+46
- FrontierMath Tiers 1-3 (v2)Claude Opus 5.5 by 12.691.278.6+12.6
- LiveBench · MathematicsClaude Opus 5.5 by 2.997.194.2+2.9
Following instructionsGPT-5.4
- LiveBench · Instruction FollowingGPT-5.4 by 4.565.770.2+4.5
Other results12 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.
- HiL-Bench (Tools-allowed)Claude Opus 5.5 by 49.3599.7+49.3
- AA-OmniscienceClaude Opus 5.5 by 40.646.45.8+40.6
- DeepSWE 1.1Claude Opus 5.5 by 22.274.252+22.2
- AA IntelligenceClaude Opus 5.5 by 18.657.639+18.6
- OfficeQA ProClaude Opus 5.5 by 16.667.751.1+16.6
- GDPval-AA (Elo)Claude Opus 5.5 by 146 rating points18201674+146 rating
- OfficeQAClaude Opus 5.5 by 10.878.968.1+10.8
- Legal Agent BenchmarkClaude Opus 5.5 by 7.98.30.4+7.9
- HealthBench length-adjustedClaude Opus 5.5 by 6.660.654+6.6
- ARC-AGI-1Claude Opus 5.5 by 3.897.593.7+3.8
- livebench_languageClaude Opus 5.5 by 3.786.382.6+3.7
- livebench_data_analysisClaude Opus 5.5 by 180.379.3+1
Questions people ask
Which is better, Claude Opus 5.5 or GPT-5.4?
Claude Opus 5.5 wins seven of the eight areas we test: coding, agents, reasoning, facts, images and charts, math and long documents. GPT-5.4 wins following instructions, and costs 27% less.
Which is better for coding?
Claude Opus 5.5. It wins 6 of the 6 coding tests both models report; GPT-5.4 wins none.
Which is cheaper?
Claude Opus 5.5 costs $4.00 per million input tokens and $20.00 per million output tokens; GPT-5.4 costs $2.50 and $15.00. That makes GPT-5.4 about 27% cheaper for the same work.
How do you compare the two?
We use the 34 benchmark tests both models have published scores on. The verdict counts the 22 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 12 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.