Claude Sonnet 4.6 vs DeepSeek-V3.2
Wins 5 of 6 areas
Coding · Agents · Reasoning · Facts · Long documents
Wins 1 of 6 areas
Following instructions
Claude Sonnet 4.6 is the stronger all-rounder.DeepSeek-V3.2 is cheaper and better at following instructions.
Scores updated · 35 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.
- CodingWriting and fixing software60Claude Sonnet 4.66 of 6 tests
- AgentsCarrying out multi-step tasks on its own40Claude Sonnet 4.64 of 5 tests · 1 tie
- ReasoningHard problems that need careful thinking30Claude Sonnet 4.63 of 4 tests · 1 tie
- FactsGetting facts right instead of making them up31Claude Sonnet 4.63 of 4 tests
- Long documentsFinding answers in very long texts10Claude Sonnet 4.61 of 1 test
- Following instructionsDoing exactly what it is asked01DeepSeek-V3.21 of 1 test
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-V3.2 costs 96% less for the same work.
The biggest differences
The tests each model wins by the widest margin, up to three each. Scores are out of 100.
Where Claude Sonnet 4.6 pulls ahead
- Abstract visual puzzles that people can solveARC-AGI-2+54.3points ahead
- Long, multi-file coding tasks in real codebasesSWE-bench Pro+42.5points ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+34.3points ahead
Where DeepSeek-V3.2 pulls ahead
- Follows unfamiliar, precisely checkable instructionsIFBench+4.1points ahead
Every test, side by side
All 35 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingClaude Sonnet 4.6
- SWE-bench ProClaude Sonnet 4.6 by 42.558.115.6+42.5
- Terminal-Bench 2.1Claude Sonnet 4.6 by 24.471.246.8+24.4
- Terminal-Bench HardClaude Sonnet 4.6 by 17.45335.6+17.4
- LMArena · WebDevClaude Sonnet 4.6 by 160 rating points15221362+160 rating
- SciCodeClaude Sonnet 4.6 by 11.250.138.9+11.2
- SWE-bench VerifiedClaude Sonnet 4.6 by 6.579.673.1+6.5
AgentsClaude Sonnet 4.6
- GDPValClaude Sonnet 4.6 by 26.936.79.8+26.9
- BrowseCompClaude Sonnet 4.6 by 23.374.751.4+23.3
- τ-Bench V3 · BankingClaude Sonnet 4.6 by 15.634.418.8+15.6
- AA ApexAgentsClaude Sonnet 4.6 by 13.52814.5+13.5
- MCP Atlastie61.362.2tie
ReasoningClaude Sonnet 4.6
- ARC-AGI-2Claude Sonnet 4.6 by 54.358.34+54.3
- Humanity's Last ExamClaude Sonnet 4.6 by 933.624.6+9
- GPQA DiamondClaude Sonnet 4.6 by 3.587.584+3.5
- CritPttie3.12.9tie
FactsClaude Sonnet 4.6
- AA-Omniscience · Non-hallucinationClaude Sonnet 4.6 by 34.351.617.3+34.3
- AA-Omniscience · AccuracyClaude Sonnet 4.6 by 7.940.933+7.9
- SimpleQA VerifiedClaude Sonnet 4.6 by 5.332.827.5+5.3
- Vectara HHEM hallucination ratelower is betterDeepSeek-V3.2 by 4.310.66.3+4.3
Following instructionsDeepSeek-V3.2
- IFBenchDeepSeek-V3.2 by 4.156.660.7+4.1
Other results14 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.
- AA-OmniscienceClaude Sonnet 4.6 by 34.712.2-22.5+34.7
- ARC-AGI-1Claude Sonnet 4.6 by 298657+29
- Artificial Analysis Coding IndexClaude Sonnet 4.6 by 18.86344.2+18.8
- τ²-Bench Telecom (AA run)DeepSeek-V3.2 by 14.975.790.6+14.9
- AA Agentic IndexClaude Sonnet 4.6 by 14.833.118.3+14.8
- Terminal-Bench 2.0Claude Sonnet 4.6 by 12.759.146.4+12.7
- GPQA (unspecified)Claude Sonnet 4.6 by 1089.979.9+10
- AA IntelligenceClaude Sonnet 4.6 by 8.630.121.5+8.6
- vectara_answer_rateClaude Sonnet 4.6 by 7.399.992.6+7.3
- frontiermath_tier_4_v1Claude Sonnet 4.6 by 6.28.32.1+6.2
- HLE (with tools)Claude Sonnet 4.6 by 646.840.8+6
- ToolathlonClaude Sonnet 4.6 by 5.84135.2+5.8
- vectara_avg_summary_lengthClaude Sonnet 4.6 by 52.7 rating points114.762+52.7 rating
- vectara_factual_consistencyDeepSeek-V3.2 by 4.389.493.7+4.3
Questions people ask
Which is better, Claude Sonnet 4.6 or DeepSeek-V3.2?
Claude Sonnet 4.6 wins five of the six areas where both have results: coding, agents, reasoning, facts and long documents. DeepSeek-V3.2 wins following instructions, and costs 96% less.
Which is better for coding?
Claude Sonnet 4.6. It wins 6 of the 6 coding tests both models report; DeepSeek-V3.2 wins none.
Which is cheaper?
Claude Sonnet 4.6 costs $3.00 per million input tokens and $15.00 per million output tokens; DeepSeek-V3.2 costs $0.28 and $0.42. That makes DeepSeek-V3.2 about 96% cheaper for the same work.
How do you compare the two?
We use the 35 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 14 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.