Claude Opus 4.6 vs DeepSeek-V4.1-Flash
Wins 2 of 8 areas
Facts · Images and charts
Wins 5 of 8 areas
Coding · Reasoning · Math · Long documents · Following instructions
DeepSeek-V4.1-Flash is the stronger all-rounder.Claude Opus 4.6 is better at facts.
Scores updated · 38 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 software03DeepSeek-V4.1-Flash3 of 4 tests · 1 tie
- MathCompetition and research-level math03DeepSeek-V4.1-Flash3 of 3 tests
- ReasoningHard problems that need careful thinking13DeepSeek-V4.1-Flash3 of 7 tests · 3 ties
- Long documentsFinding answers in very long texts01DeepSeek-V4.1-Flash1 of 1 test
- Following instructionsDoing exactly what it is asked01DeepSeek-V4.1-Flash1 of 1 test
- FactsGetting facts right instead of making them up20Claude Opus 4.62 of 3 tests · 1 tie
- Images and chartsUnderstanding pictures, charts and video21Claude Opus 4.62 of 3 tests
- AgentsCarrying out multi-step tasks on its own00Even0 each · 1 tie
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.
DeepSeek-V4.1-Flash costs 95% less for the same work.
The biggest differences
The three tests each model wins by the widest margin. Scores are out of 100.
Where DeepSeek-V4.1-Flash pulls ahead
- Fixes real GitHub issues, working as an agentLiveBench · Agentic Coding+28.3points ahead
- Extremely hard research-level math problemsFrontierMath Tier 4+19.5points ahead
- Abstract visual puzzles that people can solveARC-AGI-2+9.3points ahead
Where Claude Opus 4.6 pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+33.7points ahead
- Reasoning about charts from research papersCharXiv (RQ)+8.4points ahead
- Short factual questions, answered correctlySimpleQA Verified+4.1points ahead
Every test, side by side
All 38 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingDeepSeek-V4.1-Flash
- LiveBench · Agentic CodingDeepSeek-V4.1-Flash by 28.34977.3+28.3
- LMArena · WebDevDeepSeek-V4.1-Flash by 73 rating points15461619+73 rating
- LiveBench · CodingDeepSeek-V4.1-Flash by 1.878.280+1.8
- SciCodetie51.951.9tie
ReasoningDeepSeek-V4.1-Flash
- ARC-AGI-2DeepSeek-V4.1-Flash by 9.368.878.1+9.3
- LiveBench · ReasoningClaude Opus 4.6 by 288.786.7+2
- CritPtDeepSeek-V4.1-Flash by 1.712.614.3+1.7
- GPQA DiamondDeepSeek-V4.1-Flash by 1.389.690.9+1.3
- SimpleBenchtie67.666.7tie
- Humanity's Last Examtie39.939.2tie
- ARC-AGI-3tie0.50.2tie
FactsClaude Opus 4.6
- AA-Omniscience · Non-hallucinationClaude Opus 4.6 by 33.737.23.5+33.7
- SimpleQA VerifiedClaude Opus 4.6 by 4.14742.9+4.1
- AA-Omniscience · Accuracytie4746.4tie
Images and chartsClaude Opus 4.6
- CharXiv (RQ)Claude Opus 4.6 by 8.477.469+8.4
- LMArena · VisionClaude Opus 4.6 by 38 rating points13161278+38 rating
- MMMU-ProDeepSeek-V4.1-Flash by 1.675.477+1.6
MathDeepSeek-V4.1-Flash
- FrontierMath Tier 4DeepSeek-V4.1-Flash by 19.526.846.3+19.5
- FrontierMath Tiers 1-3 (v2)DeepSeek-V4.1-Flash by 5.26671.2+5.2
- LiveBench · MathematicsDeepSeek-V4.1-Flash by 489.393.3+4
Following instructionsDeepSeek-V4.1-Flash
- LiveBench · Instruction FollowingDeepSeek-V4.1-Flash by 6.763.370+6.7
Other results15 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.
- BabyVisionDeepSeek-V4.1-Flash by 74.814.889.6+74.8
- baby_vision_with_pythonDeepSeek-V4.1-Flash by 54.138.492.5+54.1
- AA-OmniscienceClaude Opus 4.6 by 1913.7-5.3+19
- CyberGymDeepSeek-V4.1-Flash by 14.373.888.1+14.3
- NL2RepoDeepSeek-V4.1-Flash by 14.249.864+14.2
- HLE (with tools)DeepSeek-V4.1-Flash by 10.653.363.9+10.6
- livebench_data_analysisDeepSeek-V4.1-Flash by 9.469.979.3+9.4
- vectara_factual_consistencyDeepSeek-V4.1-Flash by 887.895.8+8
- AA IntelligenceDeepSeek-V4.1-Flash by 7.631.939.5+7.6
- mmmu_pro_with_pythonDeepSeek-V4.1-Flash by 5.277.382.5+5.2
- vectara_avg_summary_lengthClaude Opus 4.6 by 29.4 rating points137.6108.2+29.4 rating
- livebench_languageClaude Opus 4.6 by 2.183.381.2+2.1
- charxiv_rq_with_pythontie84.785.6tie
- ARC-AGI-1tie9393.5tie
- vectara_answer_ratetie99.899.9tie
Questions people ask
Which is better, Claude Opus 4.6 or DeepSeek-V4.1-Flash?
DeepSeek-V4.1-Flash wins five of the eight areas we test: coding, reasoning, math, long documents and following instructions. Claude Opus 4.6 wins facts and images and charts. They are level on agents.
Which is better for coding?
DeepSeek-V4.1-Flash. It wins 3 of the 4 coding tests both models report; Claude Opus 4.6 wins none, and 1 is a tie.
Which is cheaper?
Claude Opus 4.6 costs $5.00 per million input tokens and $25.00 per million output tokens; DeepSeek-V4.1-Flash costs $0.30 and $1.20. That makes DeepSeek-V4.1-Flash about 95% cheaper for the same work.
How do you compare the two?
We use the 38 benchmark tests both models have published scores on. The verdict counts the 23 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 15 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.