Claude Sonnet 4.5 vs DeepSeek-V3
Wins 6 of 6 areas
Coding · Agents · Reasoning · Facts · Long documents · Following instructions
Wins 0 of 6 areas
—
Claude Sonnet 4.5 is the stronger all-rounder.DeepSeek-V3 is cheaper.
Scores updated · 41 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 software50Claude Sonnet 4.55 of 5 tests
- ReasoningHard problems that need careful thinking40Claude Sonnet 4.54 of 4 tests
- AgentsCarrying out multi-step tasks on its own20Claude Sonnet 4.52 of 3 tests · 1 tie
- Long documentsFinding answers in very long texts20Claude Sonnet 4.52 of 2 tests
- Following instructionsDoing exactly what it is asked20Claude Sonnet 4.52 of 2 tests
- FactsGetting facts right instead of making them up21Claude Sonnet 4.52 of 3 tests
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 costs 94% 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 4.5 pulls ahead
- Command-line tasks in a real terminalTerminal-Bench 2.1+38.9points ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+36.8points ahead
- Common-sense trick questionsSimpleBench+35.4points ahead
Where DeepSeek-V3 pulls ahead
No clear win on a test scored out of 100.
Every test, side by side
All 41 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingClaude Sonnet 4.5
- Terminal-Bench 2.1Claude Sonnet 4.5 by 38.955.816.9+38.9
- SWE-bench VerifiedClaude Sonnet 4.5 by 35.277.242+35.2
- Terminal-Bench HardClaude Sonnet 4.5 by 20.435.615.2+20.4
- LiveCodeBench v6Claude Sonnet 4.5 by 17.16446.9+17.1
- SciCodeClaude Sonnet 4.5 by 6.745.739+6.7
AgentsClaude Sonnet 4.5
- GDPValClaude Sonnet 4.5 by 20.520.50+20.5
- τ-Bench V3 · BankingClaude Sonnet 4.5 by 19.824.54.7+19.8
- Terminal-Bench 4.0tie00tie
ReasoningClaude Sonnet 4.5
- SimpleBenchClaude Sonnet 4.5 by 35.454.318.9+35.4
- GPQA DiamondClaude Sonnet 4.5 by 17.983.465.5+17.9
- Humanity's Last ExamClaude Sonnet 4.5 by 13.117.84.7+13.1
- CritPtClaude Sonnet 4.5 by 1.11.10+1.1
FactsClaude Sonnet 4.5
- AA-Omniscience · Non-hallucinationClaude Sonnet 4.5 by 36.850.914.1+36.8
- AA-Omniscience · AccuracyClaude Sonnet 4.5 by 7.432.925.4+7.4
- Vectara HHEM hallucination ratelower is betterDeepSeek-V3 by 5.9126.1+5.9
Long documentsClaude Sonnet 4.5
- AA-LCRClaude Sonnet 4.5 by 31.672.340.7+31.6
- LongBench v2Claude Sonnet 4.5 by 13.161.848.7+13.1
Following instructionsClaude Sonnet 4.5
- Multi-ChallengeClaude Sonnet 4.5 by 23.955.331.4+23.9
- IFBenchClaude Sonnet 4.5 by 16.357.341+16.3
Other results22 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.
- τ²-BenchClaude Sonnet 4.5 by 65.59832.5+65.5
- AIR-Bench 2024Claude Sonnet 4.5 by 4989.840.8+49
- HMMT 2025Claude Sonnet 4.5 by 47.174.627.5+47.1
- HarmBenchClaude Sonnet 4.5 by 42.39249.7+42.3
- AA-OmniscienceClaude Sonnet 4.5 by 40.6-0.1-40.7+40.6
- HMMT Feb. 2025Claude Sonnet 4.5 by 38.367.529.2+38.3
- AIME 2025Claude Sonnet 4.5 by 35.78751.3+35.7
- τ²-Bench Telecom (AA run)Claude Sonnet 4.5 by 3178.147.1+31
- Artificial Analysis Coding IndexClaude Sonnet 4.5 by 29.152.123+29.1
- LiveCodeBenchClaude Sonnet 4.5 by 21.87149.2+21.8
- τ²-Bench (Retail)Claude Sonnet 4.5 by 17.186.269.1+17.1
- MMLU-ProClaude Sonnet 4.5 by 11.687.575.9+11.6
- AA IntelligenceClaude Sonnet 4.5 by 1120.79.7+11
- MMMLUClaude Sonnet 4.5 by 9.789.179.4+9.7
- MMLU-ReduxClaude Sonnet 4.5 by 6.595.689.1+6.5
- vectara_factual_consistencyDeepSeek-V3 by 5.98893.9+5.9
- simple_safety_testsClaude Sonnet 4.5 by 4.710095.3+4.7
- vectara_avg_summary_lengthClaude Sonnet 4.5 by 46.1 rating points127.881.7+46.1 rating
- bbqClaude Sonnet 4.5 by 2.298.996.7+2.2
- vectara_answer_rateDeepSeek-V3 by 1.995.697.5+1.9
- anthropic_red_teamClaude Sonnet 4.5 by 1.198.297.1+1.1
- XSTesttie96.497.1tie
Questions people ask
Which is better, Claude Sonnet 4.5 or DeepSeek-V3?
Claude Sonnet 4.5 wins all six areas where both have results: coding, agents, reasoning, facts, long documents and following instructions. DeepSeek-V3 wins none, but costs 94% less.
Which is better for coding?
Claude Sonnet 4.5. It wins 5 of the 5 coding tests both models report; DeepSeek-V3 wins none.
Which is cheaper?
Claude Sonnet 4.5 costs $3.00 per million input tokens and $15.00 per million output tokens; DeepSeek-V3 costs $0.24 and $0.90. That makes DeepSeek-V3 about 94% cheaper for the same work.
How do you compare the two?
We use the 41 benchmark tests both models have published scores on. The verdict counts the 19 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 22 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.