Claude Opus 4.6 vs Claude Sonnet 4
Wins 5 of 7 areas
Coding · Agents · Reasoning · Images and charts · Long documents
Wins 2 of 7 areas
Facts · Following instructions
Claude Opus 4.6 is the stronger all-rounder.Claude Sonnet 4 is cheaper and better at following instructions.
Scores updated · 39 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 software60Claude Opus 4.66 of 6 tests
- ReasoningHard problems that need careful thinking50Claude Opus 4.65 of 5 tests
- AgentsCarrying out multi-step tasks on its own30Claude Opus 4.63 of 3 tests
- Images and chartsUnderstanding pictures, charts and video20Claude Opus 4.62 of 2 tests
- Long documentsFinding answers in very long texts10Claude Opus 4.61 of 1 test
- Following instructionsDoing exactly what it is asked02Claude Sonnet 42 of 2 tests
- FactsGetting facts right instead of making them up12Claude Sonnet 42 of 3 tests
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.
Claude Sonnet 4 costs 40% 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+71.8points ahead
- Abstract visual puzzles that people can solveARC-AGI-2+62.9points ahead
- Real work tasks from 44 professionsGDPVal+46.2points ahead
Where Claude Sonnet 4 pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+33.7points ahead
- Keeps track of context across a multi-turn chatMulti-Challenge+19.9points ahead
- Follows unfamiliar, precisely checkable instructionsIFBench+1.6points ahead
Every test, side by side
All 39 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingClaude Opus 4.6
- LiveCodeBench v6Claude Opus 4.6 by 40.388.848.5+40.3
- SWE-bench MultilingualClaude Opus 4.6 by 20.977.856.9+20.9
- Terminal-Bench HardClaude Opus 4.6 by 15.146.231.1+15.1
- SciCodeClaude Opus 4.6 by 11.951.940+11.9
- SWE-bench ProClaude Opus 4.6 by 10.753.442.7+10.7
- SWE-bench VerifiedClaude Opus 4.6 by 8.180.872.7+8.1
AgentsClaude Opus 4.6
- BrowseCompClaude Opus 4.6 by 71.88412.2+71.8
- GDPValClaude Opus 4.6 by 46.255.99.7+46.2
- OSWorld-VerifiedClaude Opus 4.6 by 30.572.742.2+30.5
ReasoningClaude Opus 4.6
- ARC-AGI-2Claude Opus 4.6 by 62.968.85.9+62.9
- Humanity's Last ExamClaude Opus 4.6 by 29.239.910.7+29.2
- SimpleBenchClaude Opus 4.6 by 22.167.645.5+22.1
- CritPtClaude Opus 4.6 by 12.312.60.3+12.3
- GPQA DiamondClaude Opus 4.6 by 11.989.677.7+11.9
FactsClaude Sonnet 4
- AA-Omniscience · Non-hallucinationClaude Sonnet 4 by 33.737.270.9+33.7
- AA-Omniscience · AccuracyClaude Opus 4.6 by 24.34722.7+24.3
- Vectara HHEM hallucination ratelower is betterClaude Sonnet 4 by 1.912.210.3+1.9
Images and chartsClaude Opus 4.6
- MMMU-ProClaude Opus 4.6 by 13.675.461.8+13.6
- LMArena · VisionClaude Opus 4.6 by 125 rating points13161191+125 rating
Following instructionsClaude Sonnet 4
- Multi-ChallengeClaude Sonnet 4 by 19.937.157.1+19.9
- IFBenchClaude Sonnet 4 by 1.653.154.7+1.6
Other results17 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.
- ARC-AGI-1Claude Opus 4.6 by 539340+53
- AA Agentic IndexClaude Opus 4.6 by 5067.617.6+50
- HLE (with tools)Claude Opus 4.6 by 3353.320.3+33
- AIME 2025Claude Opus 4.6 by 29.399.870.5+29.3
- τ²-Bench Telecom (AA run)Claude Opus 4.6 by 27.592.164.6+27.5
- frontiermath_tier_4_v1Claude Opus 4.6 by 22.922.90+22.9
- LiveCodeBenchClaude Opus 4.6 by 20.388.868.5+20.3
- τ²-Bench (Retail)Claude Opus 4.6 by 16.991.975+16.9
- AA-OmniscienceClaude Opus 4.6 by 13.513.70.2+13.5
- AA IntelligenceClaude Opus 4.6 by 1331.918.9+13
- Artificial Analysis Coding IndexClaude Opus 4.6 by 10.548.137.6+10.5
- MMLU-ProClaude Opus 4.6 by 5.489.183.7+5.4
- MMMLUClaude Opus 4.6 by 4.691.186.5+4.6
- vectara_factual_consistencyClaude Sonnet 4 by 1.987.889.7+1.9
- LiveBenchClaude Opus 4.6 by 1.576.374.8+1.5
- vectara_answer_rateClaude Opus 4.6 by 1.299.898.6+1.2
- vectara_avg_summary_lengthtie137.6145.8tie
Questions people ask
Which is better, Claude Opus 4.6 or Claude Sonnet 4?
Claude Opus 4.6 wins five of the seven areas where both have results: coding, agents, reasoning, images and charts and long documents. Claude Sonnet 4 wins facts and following instructions, and costs 40% less.
Which is better for coding?
Claude Opus 4.6. It wins 6 of the 6 coding tests both models report; Claude Sonnet 4 wins none.
Which is cheaper?
Claude Opus 4.6 costs $5.00 per million input tokens and $25.00 per million output tokens; Claude Sonnet 4 costs $3.00 and $15.00. That makes Claude Sonnet 4 about 40% cheaper for the same work.
How do you compare the two?
We use the 39 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 17 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.