Claude Opus 4.6 vs Claude Opus 5
Wins 0 of 8 areas
—
Wins 7 of 8 areas
Coding · Agents · Reasoning · Facts · Images and charts · Math · Long documents
Claude Opus 5 is the stronger all-rounder.
Scores updated · 45 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 software07Claude Opus 57 of 7 tests
- ReasoningHard problems that need careful thinking07Claude Opus 57 of 7 tests
- AgentsCarrying out multi-step tasks on its own04Claude Opus 54 of 4 tests
- FactsGetting facts right instead of making them up03Claude Opus 53 of 3 tests
- MathCompetition and research-level math03Claude Opus 53 of 3 tests
- Images and chartsUnderstanding pictures, charts and video02Claude Opus 52 of 3 tests · 1 tie
- Long documentsFinding answers in very long texts01Claude Opus 51 of 1 test
- Following instructionsDoing exactly what it is asked00Even0 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.
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 pulls ahead
- Extremely hard research-level math problemsFrontierMath Tier 4+46.4points ahead
- Solves new puzzle games it has never seenARC-AGI-3+29.7points ahead
- Long, multi-file coding tasks in real codebasesSWE-bench Pro+25.8points ahead
Where Claude Opus 4.6 pulls ahead
No clear win on a test scored out of 100.
Every test, side by side
All 45 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingClaude Opus 5
- SWE-bench ProClaude Opus 5 by 25.853.479.2+25.8
- LiveBench · Agentic CodingClaude Opus 5 by 16.24965.2+16.2
- SWE-bench VerifiedClaude Opus 5 by 15.280.896+15.2
- LMArena · WebDevClaude Opus 5 by 146 rating points15461692+146 rating
- SWE-bench MultilingualClaude Opus 5 by 11.777.889.5+11.7
- SciCodeClaude Opus 5 by 4.551.956.4+4.5
- LiveBench · CodingClaude Opus 5 by 3.378.281.5+3.3
AgentsClaude Opus 5
- MCP AtlasClaude Opus 5 by 23.162.785.8+23.1
- OSWorld-VerifiedClaude Opus 5 by 10.772.783.4+10.7
- BrowseCompClaude Opus 5 by 6.88490.8+6.8
- GDPValClaude Opus 5 by 5.355.961.2+5.3
ReasoningClaude Opus 5
- ARC-AGI-3Claude Opus 5 by 29.70.530.2+29.7
- ARC-AGI-2Claude Opus 5 by 21.668.890.4+21.6
- CritPtClaude Opus 5 by 16.512.629.1+16.5
- Humanity's Last ExamClaude Opus 5 by 1539.954.9+15
- SimpleBenchClaude Opus 5 by 1367.680.6+13
- GPQA DiamondClaude Opus 5 by 3.689.693.2+3.6
- LiveBench · ReasoningClaude Opus 5 by 2.588.791.2+2.5
FactsClaude Opus 5
- AA-Omniscience · AccuracyClaude Opus 5 by 13.94760.9+13.9
- SimpleQA VerifiedClaude Opus 5 by 12.94759.9+12.9
- AA-Omniscience · Non-hallucinationClaude Opus 5 by 237.239.2+2
Images and chartsClaude Opus 5
- MMMU-ProClaude Opus 5 by 9.375.484.7+9.3
- CharXiv (RQ)Claude Opus 5 by 6.377.483.7+6.3
- LMArena · Visiontie13161321tie
MathClaude Opus 5
- FrontierMath Tier 4Claude Opus 5 by 46.426.873.2+46.4
- FrontierMath Tiers 1-3 (v2)Claude Opus 5 by 19.66685.6+19.6
- LiveBench · MathematicsClaude Opus 5 by 6.489.395.7+6.4
Following instructionsEven
- LiveBench · Instruction Followingtie63.363.8tie
Other results16 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.
- SWE-bench MultimodalClaude Opus 5 by 32.327.159.4+32.3
- Artificial Analysis Coding IndexClaude Opus 5 by 29.948.178+29.9
- JobBenchClaude Opus 5 by 29.136.665.7+29.1
- AA-OmniscienceClaude Opus 5 by 23.413.737.1+23.4
- AA IntelligenceClaude Opus 5 by 18.931.950.8+18.9
- HiL-Bench (Tools-allowed)Claude Opus 5 by 18.738.357+18.7
- LVBenchClaude Opus 5 by 12.46375.4+12.4
- AA Agentic IndexClaude Opus 4.6 by 11.467.656.2+11.4
- HLE (with tools)Claude Opus 5 by 10.353.363.6+10.3
- OfficeQA ProClaude Opus 5 by 9.857.166.9+9.8
- GDPval-AA (Elo)Claude Opus 5 by 89 rating points16191708+89 rating
- Legal Agent BenchmarkClaude Opus 5 by 7.54.211.7+7.5
- livebench_languageClaude Opus 5 by 5.483.388.7+5.4
- livebench_data_analysisClaude Opus 5 by 4.769.974.5+4.7
- OfficeQAClaude Opus 5 by 4.673.578.1+4.6
- ARC-AGI-1Claude Opus 5 by 4.59397.5+4.5
Questions people ask
Which is better, Claude Opus 4.6 or Claude Opus 5?
Claude Opus 5 wins seven of the eight areas we test: coding, agents, reasoning, facts, images and charts, math and long documents. Claude Opus 4.6 wins none. They are level on following instructions.
Which is better for coding?
Claude Opus 5. It wins 7 of the 7 coding tests both models report; Claude Opus 4.6 wins none.
Which is cheaper?
Claude Opus 4.6 costs $5.00 per million input tokens and $25.00 per million output tokens; Claude Opus 5 costs $5.00 and $25.00. For a million tokens read plus a million written, they cost about the same.
How do you compare the two?
We use the 45 benchmark tests both models have published scores on. The verdict counts the 29 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 16 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.