Claude Sonnet 5.5 vs DeepSeek V4 Pro
Claude Sonnet 5.5 
DeepSeek V4 Pro Verdict
- Claude Sonnet 5.5 (max) leads on quality: 67.6 vs 62.8.
- Claude Sonnet 5.5 (max) is stronger in agents & tools, coding, composite, knowledge, long context, maths, reasoning.
- DeepSeek V4 Pro (max) is stronger in instruction following.
- DeepSeek V4 Pro (max) is 2.0× cheaper ($1.98 vs $4.00 per 1M blended).
- Claude Sonnet 5.5 (max) streams 1.5× faster (141 vs 93 tokens per second).
| Metric | Claude Sonnet 5.5 (max) | DeepSeek V4 Pro (max) |
|---|---|---|
| BenchLeader Index | 67.6 | 62.8 |
| Agents & tools score | 69.5 | 63.7 |
| Coding score | 69.8 | 58.3 |
| Composite score | 72.6 | 70.3 |
| Knowledge score | 65.4 | 59.8 |
| Long context score | 66.6 | 65.4 |
| Maths score | 72.9 | 56.3 |
| Reasoning score | 81.1 | 65.2 |
| Instruction following score | – | 74.3 |
| Blended price $/M | $4.00 | $1.98 |
| Output speed | 141 tok/s | 93 tok/s |
| Time to first answer | 487.1 s | 48.9 s |
| Context window | 1M | 1M |
| GPQA Diamond | 95.6% | 91.7% |
| FrontierMath Tiers 1–3 | 88.8% | 64.6% |
| FrontierMath Tier 4 | 80.5% | 26.8% |
| OTIS Mock AIME | 100.0% | 98.6% |
| SWE-bench Verified (Epoch) | – | 77.6% |
| SimpleQA Verified | 46.5% | 52.9% |
| Terminal-Bench | 61.8% | – |
| SciCode | 61.0% | 51.0% |
| WeirdML | – | 66.2% |
| APEX-Agents | 75.5% | – |
| ProofBench | 100.0% | 16.0% |
| LMArena Agent | 12 | – |
| LiveBench | 75.7% | – |
| LiveBench Reasoning | 91.6% | – |
| LiveBench Coding | 91.4% | – |
| LiveBench Agentic Coding | 56.3% | – |
| LiveBench Mathematics | 96.1% | – |
| LiveBench Data Analysis | 59.5% | – |
| LiveBench Language | 78.0% | – |
| LiveBench Instruction Following | 56.8% | – |
| AA Intelligence Index v4.3.2 | 56 | 36 |
| IFBench | – | 76.5% |
| AA-LCR | 82.7% | 80.3% |
| AA-Omniscience | 32.3 | 0.8 |
| Terminal-Bench Hard | – | 46.2% |
| GPQA Diamond (AA) | – | 92.8% |
| Humanity's Last Exam (AA) | 55.0% | 41.0% |
| SciCode (AA) | 61.0% | 51.0% |
| τ²-Bench Telecom (AA) | – | 96.2% |
| LiveCodeBench | – | 87.5% |
| MMLU-Pro | – | 87.3% |
| IOI | – | 51.6% |
| LegalBench | – | 82.4% |
| CorpFin | – | 65.4% |
| TaxEval | – | 73.1% |
| Terminal-Bench 2.1 (Vals) | – | 54.7% |
| SWE-bench (Vals) | – | 96.4% |
| GPQA Diamond (Vals) | – | 92.4% |
| Vals Index | – | 47.6 |
| AIME 2026 | – | 96.7% |
| HMMT February 2026 | – | 93.9% |
| MathArena Apex | – | 28.1% |
| ARC-AGI-1 | – | 90.0% |
| ARC-AGI-2 | – | 61.3% |
| CritPt | 31.4% | 18.0% |
| GDPval-AA v2.1 | 67.0% | 47.5% |
| τ³-Banking (AA) | – | 39.6% |
| ITBench SRE (AA) | – | 38.3% |
| Analyst Agent (AA) | 57.5% | 18.8% |
| APEX-Agents (AA) | – | 24.3% |
| CaseLaw v2 | – | 59.4% |
| Code Migration | – | 41.5% |
| CyberBench | – | 72.9% |
| Excel Modeling Benchmark | – | 52.8% |
| Finance Agent v2 | – | 50.4% |
| Harvey's Legal Agent Benchmark | – | 7.5% |
| Legal Research Bench | – | 40.9% |
| MedCode | – | 42.5% |
| MedScribe | – | 80.2% |
| ProgramBench | – | 0.0% |
| Public Benefits Bench | – | 62.9% |
| SkillsBench | – | 53.8% |
| Tax Agent Bench | – | 58.7% |
| Terminal-Bench 4.0 (Vals) | – | 14.1% |
| Terminal-Bench Science | – | 0.0% |
| Vibe Code Bench 1-100 | – | 17.5% |
| Vibe Code Bench v1.1 | – | 82.3% |
| Chess Puzzles | – | 47.0% |
| Mystery Game Puzzles | 65.0% | 43.0% |
| LMCA | – | 41.2% |
| DTBench | – | 90.7% |
| CursorBench | 55.5% | – |
| ALE-Bench | – | 1403.2 |
| FrontierSWE | 61.9% | – |
| Terminal-Bench 4.0 (AA) | 63.6% | 14.7% |
| Terminal-Bench 2.1 (AA) | – | 78.7% |
| AutomationBench | 71.8% | 56.7% |
| GDP.pdf | 25.8% | 11.4% |
| MLCR | 75.0% | 17.8% |
| Harvey LAB | 2.8% | – |
| EnterpriseOps-Gym | – | 49.6% |
| AA-Omniscience: accuracy | 54.0% | 49.1% |
| AA-Omniscience: non-hallucination | 53.0% | 5.9% |
| AA-Briefcase v1.1 | 1823 | 1257 |
| AA Openness Index | – | 44.4 |
Data as of 2026-10-11. Best configuration of each model; every score links to its source on the model pages.
Claude Sonnet 5.5 vs DeepSeek V4 Pro: questions
- Is Claude Sonnet 5.5 better than DeepSeek V4 Pro?
- Claude Sonnet 5.5 (max) leads on quality: 67.6 vs 62.8. The BenchLeader Index combines every independent quality benchmark; Claude Sonnet 5.5 (max) is ahead overall as of 2026-10-11, but check the category scores for your use.
- Is Claude Sonnet 5.5 better than DeepSeek V4 Pro for coding?
- Claude Sonnet 5.5 scores higher in coding (70 vs 58 on the category index, where 50 is average).
- Is Claude Sonnet 5.5 better than DeepSeek V4 Pro for agentic tasks?
- Claude Sonnet 5.5 scores higher in agentic tasks (70 vs 64 on the category index, where 50 is average).
- Which is cheaper, Claude Sonnet 5.5 or DeepSeek V4 Pro?
- DeepSeek V4 Pro is cheaper: $1.98 against $4.00 per million tokens, blended at three input tokens per output token.
- Which is faster, Claude Sonnet 5.5 or DeepSeek V4 Pro?
- Claude Sonnet 5.5 streams faster: 141 against 93 output tokens per second.
- Which has the larger context window?
- Both accept 1M tokens of context.