DeepSeek-V4.1-Flash vs Claude Sonnet 5.5
Wins 2 of 8 areas
Long documents · Following instructions
Wins 5 of 8 areas
Coding · Agents · Reasoning · Facts · Math
Claude Sonnet 5.5 is the stronger all-rounder.DeepSeek-V4.1-Flash is cheaper and better at long documents.
Scores updated · 31 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.
- ReasoningHard problems that need careful thinking06Claude Sonnet 5.56 of 6 tests
- FactsGetting facts right instead of making them up03Claude Sonnet 5.53 of 3 tests
- MathCompetition and research-level math03Claude Sonnet 5.53 of 3 tests
- CodingWriting and fixing software13Claude Sonnet 5.53 of 4 tests
- AgentsCarrying out multi-step tasks on its own02Claude Sonnet 5.52 of 2 tests
- Long documentsFinding answers in very long texts10DeepSeek-V4.1-Flash1 of 1 test
- Following instructionsDoing exactly what it is asked10DeepSeek-V4.1-Flash1 of 1 test
- Images and chartsUnderstanding pictures, charts and video00Even0 each · 1 tie
Images and charts, 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 88% 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 Sonnet 5.5 pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+49.5points ahead
- Complex command-line tasks across many fieldsTerminal-Bench 4.0+36.8points ahead
- Extremely hard research-level math problemsFrontierMath Tier 4+34.2points ahead
Where DeepSeek-V4.1-Flash pulls ahead
- Fixes real GitHub issues, working as an agentLiveBench · Agentic Coding+21points ahead
- Follows detailed instructions exactlyLiveBench · Instruction Following+13.2points ahead
- Reasons across sets of long documentsAA-LCR+1.3points ahead
Every test, side by side
All 31 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingClaude Sonnet 5.5
- LiveBench · Agentic CodingDeepSeek-V4.1-Flash by 2177.356.3+21
- LMArena · WebDevClaude Sonnet 5.5 by 151 rating points16191770+151 rating
- LiveBench · CodingClaude Sonnet 5.5 by 11.48091.4+11.4
- SciCodeClaude Sonnet 5.5 by 9.151.961+9.1
AgentsClaude Sonnet 5.5
- Terminal-Bench 4.0Claude Sonnet 5.5 by 36.826.863.6+36.8
- GDPValClaude Sonnet 5.5 by 11.95566.9+11.9
ReasoningClaude Sonnet 5.5
- ARC-AGI-3Claude Sonnet 5.5 by 24.10.224.3+24.1
- CritPtClaude Sonnet 5.5 by 17.114.331.4+17.1
- Humanity's Last ExamClaude Sonnet 5.5 by 15.839.255+15.8
- SimpleBenchClaude Sonnet 5.5 by 9.266.775.9+9.2
- ARC-AGI-2Claude Sonnet 5.5 by 5.878.183.9+5.8
- LiveBench · ReasoningClaude Sonnet 5.5 by 4.986.791.6+4.9
FactsClaude Sonnet 5.5
- AA-Omniscience · Non-hallucinationClaude Sonnet 5.5 by 49.53.553+49.5
- AA-Omniscience · AccuracyClaude Sonnet 5.5 by 7.646.454+7.6
- SimpleQA VerifiedClaude Sonnet 5.5 by 3.642.946.5+3.6
MathClaude Sonnet 5.5
- FrontierMath Tier 4Claude Sonnet 5.5 by 34.246.380.5+34.2
- FrontierMath Tiers 1-3 (v2)Claude Sonnet 5.5 by 17.671.288.8+17.6
- LiveBench · MathematicsClaude Sonnet 5.5 by 2.893.396.1+2.8
Long documentsDeepSeek-V4.1-Flash
- AA-LCRDeepSeek-V4.1-Flash by 1.38482.7+1.3
Following instructionsDeepSeek-V4.1-Flash
- LiveBench · Instruction FollowingDeepSeek-V4.1-Flash by 13.27056.8+13.2
Other results10 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.
- Program BenchClaude Sonnet 5.5 by 59.420.379.7+59.4
- Terminal-Bench-Science 0.1Claude Sonnet 5.5 by 44.215.759.9+44.2
- Terminal-Bench 4.0Claude Sonnet 5.5 by 39.431.270.6+39.4
- AA-OmniscienceClaude Sonnet 5.5 by 37.6-5.332.3+37.6
- livebench_data_analysisDeepSeek-V4.1-Flash by 19.879.359.5+19.8
- AA IntelligenceClaude Sonnet 5.5 by 16.539.556+16.5
- DeepSWE 1.1DeepSeek-V4.1-Flash by 3.274.271+3.2
- livebench_languageDeepSeek-V4.1-Flash by 3.281.278+3.2
- ARC-AGI-1Claude Sonnet 5.5 by 393.596.5+3
- HLE (with tools)tie63.964.5tie
Questions people ask
Which is better, DeepSeek-V4.1-Flash or Claude Sonnet 5.5?
Claude Sonnet 5.5 wins five of the eight areas we test: coding, agents, reasoning, facts and math. DeepSeek-V4.1-Flash wins long documents and following instructions, and costs 88% less. They are level on images and charts.
Which is better for coding?
Claude Sonnet 5.5. It wins 3 of the 4 coding tests both models report; DeepSeek-V4.1-Flash wins 1.
Which is cheaper?
DeepSeek-V4.1-Flash costs $0.30 per million input tokens and $1.20 per million output tokens; Claude Sonnet 5.5 costs $2.00 and $10.00. That makes DeepSeek-V4.1-Flash about 88% cheaper for the same work.
How do you compare the two?
We use the 31 benchmark tests both models have published scores on. The verdict counts the 21 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 10 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.