Claude Opus 4.6 vs Qwen3.6 Max Preview
Wins 3 of 6 areas
Coding · Agents · Reasoning
Wins 3 of 6 areas
Facts · Long documents · Following instructions
The two are evenly matched.Claude Opus 4.6 is better at reasoning and coding; Qwen3.6 Max Preview at facts and long documents.
Scores updated · 19 tests both models report · How we compare
Where each one wins
Tests won in each of the six 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.
- ReasoningHard problems that need careful thinking30Claude Opus 4.63 of 4 tests · 1 tie
- CodingWriting and fixing software30Claude Opus 4.63 of 3 tests
- AgentsCarrying out multi-step tasks on its own10Claude Opus 4.61 of 1 test
- FactsGetting facts right instead of making them up12Qwen3.6 Max Preview2 of 3 tests
- Long documentsFinding answers in very long texts01Qwen3.6 Max Preview1 of 1 test
- Following instructionsDoing exactly what it is asked01Qwen3.6 Max Preview1 of 1 test
Agents, 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.
Qwen3.6 Max Preview costs 70% 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
- Very hard expert questions across many subjectsHumanity's Last Exam+9.1points ahead
- Answers hard knowledge questions correctlyAA-Omniscience · Accuracy+9.1points ahead
- Research-level physics problemsCritPt+8.9points ahead
Where Qwen3.6 Max Preview pulls ahead
- Follows unfamiliar, precisely checkable instructionsIFBench+23.5points ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+16.6points ahead
- Short factual questions, answered correctlySimpleQA Verified+5points ahead
Every test, side by side
All 19 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 64 rating points15461482+64 rating
- SciCodeClaude Opus 4.6 by 551.946.9+5
- Terminal-Bench HardClaude Opus 4.6 by 2.346.243.9+2.3
ReasoningClaude Opus 4.6
- Humanity's Last ExamClaude Opus 4.6 by 9.139.930.8+9.1
- CritPtClaude Opus 4.6 by 8.912.63.7+8.9
- SimpleBenchClaude Opus 4.6 by 4.667.663+4.6
- GPQA Diamondtie89.688.8tie
FactsQwen3.6 Max Preview
- AA-Omniscience · Non-hallucinationQwen3.6 Max Preview by 16.637.253.8+16.6
- AA-Omniscience · AccuracyClaude Opus 4.6 by 9.14737.9+9.1
- SimpleQA VerifiedQwen3.6 Max Preview by 54752+5
Long documentsQwen3.6 Max Preview
- AA-LCRQwen3.6 Max Preview by 2.77880.7+2.7
Following instructionsQwen3.6 Max Preview
- IFBenchQwen3.6 Max Preview by 23.553.176.6+23.5
Other results6 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.
- frontiermath_tier_4_v1Claude Opus 4.6 by 18.722.94.2+18.7
- AA-OmniscienceClaude Opus 4.6 by 4.513.79.2+4.5
- τ²-Bench Telecom (AA run)Qwen3.6 Max Preview by 3.892.195.9+3.8
- AA IntelligenceClaude Opus 4.6 by 3.531.928.4+3.5
- Artificial Analysis Coding IndexClaude Opus 4.6 by 3.248.144.9+3.2
- AA Agentic IndexClaude Opus 4.6 by 2.867.664.8+2.8
Questions people ask
Which is better, Claude Opus 4.6 or Qwen3.6 Max Preview?
Claude Opus 4.6 and Qwen3.6 Max Preview each win three of the six areas where both have results. Claude Opus 4.6 wins coding, agents and reasoning; Qwen3.6 Max Preview wins facts, long documents and following instructions. Qwen3.6 Max Preview costs 70% less.
Which is better for coding?
Claude Opus 4.6. It wins 3 of the 3 coding tests both models report; Qwen3.6 Max Preview wins none.
Which is cheaper?
Claude Opus 4.6 costs $5.00 per million input tokens and $25.00 per million output tokens; Qwen3.6 Max Preview costs $1.30 and $7.80. That makes Qwen3.6 Max Preview about 70% cheaper for the same work.
How do you compare the two?
We use the 19 benchmark tests both models have published scores on. The verdict counts the 13 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 6 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.