Claude Opus 4.5 vs DeepSeek-V3.1
Wins 6 of 6 areas
Coding · Agents · Reasoning · Facts · Long documents · Following instructions
Wins 0 of 6 areas
—
Claude Opus 4.5 is the stronger all-rounder.DeepSeek-V3.1 is cheaper.
Scores updated · 28 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 4.53 of 3 tests
- CodingWriting and fixing software31Claude Opus 4.53 of 4 tests
- AgentsCarrying out multi-step tasks on its own20Claude Opus 4.52 of 2 tests
- Following instructionsDoing exactly what it is asked20Claude Opus 4.52 of 2 tests
- FactsGetting facts right instead of making them up21Claude Opus 4.52 of 3 tests
- Long documentsFinding answers in very long texts10Claude Opus 4.51 of 1 test
Long documents rests on a single test.
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-V3.1 costs 93% 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 Opus 4.5 pulls ahead
- Real work tasks from 44 professionsGDPVal+41.7points ahead
- Hard command-line tasks in a real terminalTerminal-Bench Hard+22points ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+21.5points ahead
Where DeepSeek-V3.1 pulls ahead
- Fixes real GitHub issues in many programming languagesSWE-bench Multilingual+47.5points ahead
Every test, side by side
All 28 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingClaude Opus 4.5
- SWE-bench MultilingualDeepSeek-V3.1 by 47.5754.5+47.5
- Terminal-Bench HardClaude Opus 4.5 by 224725+22
- SWE-bench VerifiedClaude Opus 4.5 by 14.980.966+14.9
- SciCodeClaude Opus 4.5 by 10.449.539.1+10.4
AgentsClaude Opus 4.5
- GDPValClaude Opus 4.5 by 41.747.35.6+41.7
- BrowseCompClaude Opus 4.5 by 73730+7
ReasoningClaude Opus 4.5
- Humanity's Last ExamClaude Opus 4.5 by 15.830.114.3+15.8
- GPQA DiamondClaude Opus 4.5 by 8.786.677.9+8.7
- CritPtClaude Opus 4.5 by 2.64.62+2.6
FactsClaude Opus 4.5
- AA-Omniscience · Non-hallucinationClaude Opus 4.5 by 21.53917.5+21.5
- AA-Omniscience · AccuracyClaude Opus 4.5 by 17.646.629+17.6
- Vectara HHEM hallucination ratelower is betterDeepSeek-V3.1 by 5.410.95.5+5.4
Following instructionsClaude Opus 4.5
- IFBenchClaude Opus 4.5 by 16.55841.5+16.5
- Multi-ChallengeClaude Opus 4.5 by 12.95946.1+12.9
Other results13 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-Bench Telecom (AA run)Claude Opus 4.5 by 52.189.537.4+52.1
- AA-OmniscienceClaude Opus 4.5 by 43.614-29.6+43.6
- AIME 2025Claude Opus 4.5 by 4392.849.8+43
- AA Agentic IndexClaude Opus 4.5 by 40.759.618.9+40.7
- LiveCodeBenchClaude Opus 4.5 by 30.68756.4+30.6
- Artificial Analysis Coding IndexClaude Opus 4.5 by 18.147.829.7+18.1
- AA IntelligenceClaude Opus 4.5 by 15.629.113.5+15.6
- HMMT Feb. 2025Claude Opus 4.5 by 7.192.985.8+7.1
- MMLU-ProClaude Opus 4.5 by 5.889.583.7+5.8
- vectara_factual_consistencyDeepSeek-V3.1 by 5.489.194.5+5.4
- vectara_avg_summary_lengthClaude Opus 4.5 by 50.8 rating points114.563.7+50.8 rating
- vectara_answer_rateClaude Opus 4.5 by 4.298.794.5+4.2
- MMLU-ReduxClaude Opus 4.5 by 3.895.691.8+3.8
Questions people ask
Which is better, Claude Opus 4.5 or DeepSeek-V3.1?
Claude Opus 4.5 wins all six areas where both have results: coding, agents, reasoning, facts, long documents and following instructions. DeepSeek-V3.1 wins none, but costs 93% less.
Which is better for coding?
Claude Opus 4.5. It wins 3 of the 4 coding tests both models report; DeepSeek-V3.1 wins 1.
Which is cheaper?
Claude Opus 4.5 costs $5.00 per million input tokens and $25.00 per million output tokens; DeepSeek-V3.1 costs $0.56 and $1.68. That makes DeepSeek-V3.1 about 93% cheaper for the same work.
How do you compare the two?
We use the 28 benchmark tests both models have published scores on. The verdict counts the 15 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 13 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.