Claude Sonnet 5.5 vs DeepSeek V4 Pro 0424
Full profiles: Claude Sonnet 5.5 (Anthropic) and DeepSeek V4 Pro 0424 (DeepSeek): every benchmark, price and rank.
Updated · Data from 8 sources
As of 22 Oct 2026, Claude Sonnet 5.5 (max) leads on quality (69.4 vs 59.7 on the BenchLeader Index), but DeepSeek V4 Pro 0424 (max) costs 7.3× less, so it is the better buy unless you need the extra quality. Claude Sonnet 5.5 leads in every category we measure both in.
Claude Sonnet 5.5 benchmarks max Anthropic
DeepSeek V4 Pro 0424 benchmarks max DeepSeekOpen in the comparison toolWhich to pick for…
The pick's numbers come first. Category scores on the BenchLeader scale (50 is the average model, 15 points one standard deviation). High confidence means a lead of half a standard deviation or more; under one point is called either.
| Use case | Pick | Why | Confidence |
|---|---|---|---|
| Agents & tool use | Claude Sonnet 5.5 | 70 vs 64 category score; GDPval-AA v2.1 66.9% vs 32.8% | Medium |
| Reasoning | Claude Sonnet 5.5 | 78 vs 62 category score; CritPt 31.4% vs 12.9% | High |
| Knowledge | Claude Sonnet 5.5 | 65 vs 57 category score; AA-Omniscience 32.3 vs -10.7 | High |
| Long documents | Claude Sonnet 5.5 | 67 vs 62 category score; AA-LCR 82.7% vs 74.7% | Medium |
| Lowest cost | DeepSeek V4 Pro 0424 | $0.548 vs $4.00 per 1M blended; agentic coding session $0.030 vs $0.160 | – |
| Fastest first answer | DeepSeek V4 Pro 0424 | 381.5 s (Claude Sonnet 5.5 (max)) vs 49.0 s (DeepSeek V4 Pro 0424 (max)) to the first answer token | – |
| High volume on a budget | Neither flagship | A smaller model is usually the better fit: Cheapest good LLMs, Fastest LLMs | – |
Key differences
- Claude Sonnet 5.5 (max) leads on quality: 69.4 vs 59.7.
- Claude Sonnet 5.5 (max) is stronger in agents & tools, coding, composite, knowledge, long context, maths, reasoning.
- DeepSeek V4 Pro 0424 (max) is stronger in instruction following.
- DeepSeek V4 Pro 0424 (max) is 7.3× cheaper ($0.548 vs $4.00 per 1M blended).
- Claude Sonnet 5.5 (max) streams 1.4× faster (134 vs 92 tokens per second).
At a glance
| Spec | Claude Sonnet 5.5 | DeepSeek V4 Pro 0424 |
|---|---|---|
| Developer | Anthropic | DeepSeek |
| Released | 28 Sept 2026 | 24 Apr 2026 |
| Context window (tokens) | 1,000,000 | 1,000,000 |
| Compared at | Claude Sonnet 5.5 (max) | DeepSeek V4 Pro 0424 (max) |
| Open weights | No | Yes |
| Price in / out / cached input per 1M | $2.00 / $10.00 / $0.100 | $0.440 / $0.870 / – |
Category by category
Category scores on the BenchLeader scale: 50 is the average evaluated model, 15 points is one standard deviation. A gap under one point is called level.
| Category | Claude Sonnet 5.5 (max) | DeepSeek V4 Pro 0424 (max) | Leader |
|---|---|---|---|
| Reasoning | 77.5 | 62.2 | Claude Sonnet 5.5 |
| Coding | 69.3 | – | not comparable |
| Agents & tools | 69.7 | 63.8 | Claude Sonnet 5.5 |
| Maths | 72.4 | – | not comparable |
| Knowledge | 65.1 | 56.7 | Claude Sonnet 5.5 |
| Instruction following | – | 74.5 | not comparable |
| Long context | 66.5 | 62.3 | Claude Sonnet 5.5 |
| Composite | 92.5 | 63.7 | Claude Sonnet 5.5 |
Where sources disagree
Across the 13 benchmarks both have, Claude Sonnet 5.5 leads 13, DeepSeek V4 Pro 0424 leads 0. Every overall index that rates both puts Claude Sonnet 5.5 ahead.
| Overall index | Claude Sonnet 5.5 (max) | DeepSeek V4 Pro 0424 (max) | Leader |
|---|---|---|---|
| BenchLeader Index | 69.4 | 59.7 | Claude Sonnet 5.5 |
| AA Intelligence Index v4.3.2 | 56 | 30.4 | Claude Sonnet 5.5 |
Each publisher runs its own tasks, harnesses and effort settings, so their overall verdicts can differ; BenchLeader reconciles them on one scale. † measured at another effort setting of the same model (the publisher did not test the one compared here).
Price, speed and latency
List prices blended at three input tokens per output token. Per-task prices use the token counts on the methodology page; suite cost is what Artificial Analysis paid to run its whole Intelligence Index.
| Measure | Claude Sonnet 5.5 (max) | DeepSeek V4 Pro 0424 (max) |
|---|---|---|
| Blended price per 1M tokens | $4.00 | $0.548 |
| Input / output per 1M | $2.00 / $10.00 | $0.440 / $0.870 |
| Cached input per 1M | $0.100 | – |
| Price per task: chat reply | $0.0038 | $0.0004 |
| Price per task: summarize a 30-page report | $0.030 | $0.0058 |
| Price per task: code edit | $0.027 | $0.0039 |
| Price per task: agentic coding session | $0.160 | $0.030 |
| Price per task: structured extraction | $0.0060 | $0.0011 |
| Cost to run a full benchmark suite | $5.46 | $0.122 |
| Output speed | 134 tok/s | 92 tok/s |
| Time to first answer | 381.5 s | 49.0 s |
| Time to a complete answer | 385.2 s | 54.4 s |
Per token is not per task: on Artificial Analysis' full suite Claude Sonnet 5.5 (max) cost $5.46 and DeepSeek V4 Pro 0424 (max) $0.122, because Claude Sonnet 5.5 (max) used about 6.2× as many tokens for the same work, which a per-token price hides.
Each model is compared at its best-ranked reasoning-effort setting. Claude Sonnet 5.5 (max) scores 69.4 and takes 382 s to its first answer token; Claude Sonnet 5.5 (xhigh) scores 68.0 in 29 s, Claude Sonnet 5.5 (high) scores 65.7 in 11 s. DeepSeek V4 Pro 0424 (max) scores 59.7 and takes 49 s to its first answer token; DeepSeek V4 Pro 0424 (high) scores 57.9 in 25 s, DeepSeek V4 Pro 0424 (no reasoning) scores 50.8 in 2 s. Higher effort means more thinking before the answer: better scores, longer waits and more tokens billed. How reasoning effort changes scores and cost.
Every benchmark
102 benchmarks, grouped by category; each name links to the board and its leaders. † measured at another effort setting of the same model, shown for reference and not counted as a win.
| Metric | Publisher | Claude Sonnet 5.5 (max) | DeepSeek V4 Pro 0424 (max) |
|---|---|---|---|
| BenchLeader Index | BenchLeader | 69.4 | 59.7 |
| Agents & tools score | BenchLeader | 69.7 | 63.8 |
| Coding score | BenchLeader | 69.3 | – |
| Composite score | BenchLeader | 92.5 | 63.7 |
| Knowledge score | BenchLeader | 65.1 | 56.7 |
| Long context score | BenchLeader | 66.5 | 62.3 |
| Maths score | BenchLeader | 72.4 | – |
| Reasoning score | BenchLeader | 77.5 | 62.2 |
| Instruction following score | BenchLeader | – | 74.5 |
| Blended price $/M | BenchLeader | $4.00 | $0.548 |
| Output speed | BenchLeader | 134 tok/s | 92 tok/s |
| Time to first answer | BenchLeader | 381.5 s | 49.0 s |
| Context window | BenchLeader | 1,000,000 tokens | 1,000,000 tokens |
| Reasoning | |||
| GPQA Diamond | Epoch AI Benchmarking Hub | 95.6% | – |
| Humanity's Last Exam | Scale AI / CAIS | 46.1%† | – |
| Arena Hard Prompts | Arena (LMArena) | 1506† | – |
| LiveBench 2.0 Reasoning | LiveBench | 91.6%† | – |
| LiveBench 1.0 Reasoning | LiveBench | 91.6% | – |
| GPQA Diamond (AA) | Artificial Analysis | – | 88.8% |
| Humanity's Last Exam (AA) | Artificial Analysis | 55.0% | 37.5% |
| ARC-AGI-1 | ARC Prize | 96.5% | – |
| ARC-AGI-2 | ARC Prize | 83.9%† | – |
| ARC-AGI-3 | ARC Prize | 72.4%† | – |
| CritPt | Artificial Analysis | 31.4% | 12.9% |
| MysteryMechanism | Vals AI | 49.1%† | – |
| Chess Puzzles | Epoch AI Benchmarking Hub | 59.0% | – |
| Mystery Game Puzzles | Epoch AI Benchmarking Hub | 65.0% | – |
| LMCA | Epoch AI Benchmarking Hub | 68.3% | – |
| DTBench | Epoch AI Benchmarking Hub | 97.3% | – |
| Coding | |||
| SciCode | SciCode | 61.0% | – |
| FrontierCode | Cognition | 52.1%† | – |
| Arena Coding | Arena (LMArena) | 1530† | – |
| Arena WebDev | Arena (LMArena) | 1768† | – |
| LiveBench 2.0 Coding | LiveBench | 94.7%† | – |
| LiveBench 1.0 Coding | LiveBench | 91.4% | – |
| SciCode (AA) | Artificial Analysis | 61.0% | 50.8% |
| IOI | Vals AI | 83.1%† | – |
| Code Migration | Vals AI | 69.8%† | – |
Show all 102 benchmarks
| Metric | Publisher | Claude Sonnet 5.5 (max) | DeepSeek V4 Pro 0424 (max) |
|---|---|---|---|
| Coding | |||
| ProgramBench | Vals AI | 6.5%† | – |
| Vibe Code Bench v1.1 | Vals AI | 92.4%† | – |
| CursorBench | Cursor | 55.5% | – |
| ALE-Bench | Epoch AI Benchmarking Hub | 1819.1† | – |
| FrontierSWE | Epoch AI Benchmarking Hub | 61.9% | – |
| Terminal-Bench 4.0 (AA) | Artificial Analysis | 63.6% | 14.7% |
| Terminal-Bench 2.1 (AA) | Artificial Analysis | – | 64.0% |
| Startrise Frontend | Startrise | 75.9%† | – |
| FrontierCode 1.1 (AA) | Artificial Analysis | 45.8% | – |
| Agents & tools | |||
| Terminal-Bench | Terminal-Bench | 61.8% | – |
| APEX-Agents | Mercor | 75.5% | – |
| Arena Agent | Arena (LMArena) | 12.7 | – |
| LiveBench Long Agentic | LiveBench | 78.8%† | – |
| LiveBench Browser | LiveBench | 84.0%† | – |
| LiveBench Agentic Coding | LiveBench | 56.3% | – |
| Terminal-Bench Hard | Artificial Analysis | – | 46.2% |
| τ²-Bench Telecom (AA) | Artificial Analysis | – | 96.2% |
| Terminal-Bench 2.1 (Vals) | Vals AI | 83.2%† | – |
| GDPval-AA v2.1 | Artificial Analysis | 66.9% | 32.8% |
| τ³-Banking (AA) | Artificial Analysis | – | 30.1% |
| ITBench SRE (AA) | Artificial Analysis | – | 38.3% |
| Analyst Agent (AA) | Artificial Analysis | 57.5% | 18.8% |
| APEX-Agents (AA) | Artificial Analysis | – | 24.3% |
| CyberBench | Vals AI | 59.6%† | – |
| Finance Agent v2 | Vals AI | 58.1%† | – |
| Harvey's Legal Agent Benchmark | Vals AI | 10.4%† | – |
| Legal Research Bench | Vals AI | 48.1%† | – |
| SREBench | Vals AI | 30.1%† | – |
| Tax Agent Bench | Vals AI | 42.0%† | – |
| Terminal-Bench 4.0 (Vals) | Vals AI | 64.1%† | – |
| Terminal-Bench Science | Vals AI | 47.1%† | – |
| Vending-Bench 2 | Andon Labs | 11165.3† | – |
| AutomationBench | Artificial Analysis | 71.8% | 56.3% |
| GDP.pdf (AA) | Artificial Analysis | 25.8% | 13.4% |
| Harvey LAB-AA v1.1 | Harvey | 2.8% | – |
| EnterpriseOps-Gym | Artificial Analysis | – | 40.4% |
| AA-Briefcase v1.1 | Artificial Analysis | 1817 | – |
| Terminal-Bench-Science 0.1 (AA) | Artificial Analysis | 53.3% | – |
| Maths | |||
| FrontierMath Tiers 1–3 | Epoch AI Benchmarking Hub | 88.8% | – |
| FrontierMath Tier 4 | Epoch AI Benchmarking Hub | 80.5% | – |
| OTIS Mock AIME | Epoch AI Benchmarking Hub | 100.0% | – |
| ProofBench | Vals AI | 100.0% | – |
| LiveBench Mathematics | LiveBench | 96.1% | – |
| Arena Maths | Arena (LMArena) | 1500† | – |
| Knowledge | |||
| SimpleQA Verified | Epoch AI Benchmarking Hub | 46.5% | – |
| LiveBench 2.0 Data Analysis | LiveBench | 59.5%† | – |
| LiveBench Research | LiveBench | 83.6%† | – |
| LiveBench 1.0 Data Analysis | LiveBench | 59.5% | – |
| AA-Omniscience | Artificial Analysis | 32.3 | -10.7 |
| BioMysteryBench | Vals AI | 81.1%† | – |
| Excel Modeling Benchmark | Vals AI | 75.7%† | – |
| MedCode | Vals AI | 52.9%† | – |
| MedScribe | Vals AI | 91.1%† | – |
| Public Benefits Bench | Vals AI | 67.2%† | – |
| GDP.pdf | Epoch AI Benchmarking Hub | 31.4% | – |
| AA-Omniscience: accuracy | Artificial Analysis | 54.0% | 43.0% |
| AA-Omniscience: non-hallucination | Artificial Analysis | 53.0% | 5.9% |
| Instruction following | |||
| LiveBench Language | LiveBench | 78.0% | – |
| LiveBench Instruction Following | LiveBench | 56.8% | – |
| IFBench | Artificial Analysis | – | 76.5% |
| Arena Instruction Following | Arena (LMArena) | 1487† | – |
| Human preference | |||
| Arena Text | Arena (LMArena) | 1477† | – |
| Arena Creative Writing | Arena (LMArena) | 1465† | – |
| Arena Multi-turn | Arena (LMArena) | 1474† | – |
| Arena Longer Queries | Arena (LMArena) | 1500† | – |
| Multimodal | |||
| Arena Vision | Arena (LMArena) | 1287† | – |
| LiveBench Design | LiveBench | 85.0%† | – |
| SAGE | Vals AI | 51.8%† | – |
| Blueprint-Bench 2 | Epoch AI Benchmarking Hub | 39.3%† | – |
| AA-3D | Artificial Analysis | 1197 | – |
| Long context | |||
| AA-LCR | Artificial Analysis | 82.7% | 74.7% |
| MLCR | Artificial Analysis | 41.7% | 9.4% |
| Composite | |||
| Epoch Capabilities Index | Epoch AI Benchmarking Hub | 166.2† | – |
| LiveBench 2.0 | LiveBench | 82.5%† | – |
| LiveBench 1.0 | LiveBench | 75.7% | – |
| AA Intelligence Index v4.3.2 | Artificial Analysis | 56 | 30.4 |
| Vals Index | Vals AI | 70.8† | – |
Data as of 2026-10-22. Best configuration of each model; every score links to its source on the model pages.
What changed
- Claude Sonnet 5.5 (max): 1 new result (ARC-AGI-1); 1 score revision.
- Claude Sonnet 5.5 (max): BenchLeader Index 67.6 → 69.5.
- Claude Sonnet 5.5 (max): 2 new results (LMCA, DTBench).
- Claude Sonnet 5.5 (max): 1 score revision.
- Claude Sonnet 5.5 (max): 1 score revision.
- Claude Sonnet 5.5 (max): 2 new results (Chess Puzzles, GDP.pdf); 1 score revision.
- Claude Sonnet 5.5 (max): 1 score revision.
- Claude Sonnet 5.5 (max): 3 new results (FrontierCode 1.1 (AA), Terminal-Bench-Science 0.1 (AA), AA-3D).
- Claude Sonnet 5.5 (max): 2 score revisions.
- Claude Sonnet 5.5 (max): BenchLeader Index 61.7 → 67.6.
Related comparisons
- Claude Sonnet 5.5 vs Claude Opus 5.5
- Claude Sonnet 5.5 vs Gemini 4 Argon
- Claude Sonnet 5.5 vs Muse Wonder Full
- DeepSeek V4 Pro 0424 vs Claude Opus 5.5
- DeepSeek V4 Pro 0424 vs Gemini 4 Argon
- DeepSeek V4 Pro 0424 vs Muse Wonder Full
- DeepSeek V4 Pro 0424 vs Claude Fable 5.1
- Claude Sonnet 5.5 vs GPT-6 Astra
Rankings: Best LLM overall · Best LLM for agents · Best cheap LLMs
Claude Sonnet 5.5 vs DeepSeek V4 Pro 0424: questions
- Is Claude Sonnet 5.5 better than DeepSeek V4 Pro 0424?
- Claude Sonnet 5.5 (max) leads on quality: 69.4 vs 59.7. The BenchLeader Index combines every independent quality benchmark; Claude Sonnet 5.5 (max) is ahead overall as of 2026-10-22, but check the category scores for your use.
- Which is cheaper, Claude Sonnet 5.5 or DeepSeek V4 Pro 0424, and will that change?
- DeepSeek V4 Pro 0424 is cheaper today: $0.548 against $4.00 per million tokens, blended at three input tokens per output token. No price change has been announced for either.
- Is Claude Sonnet 5.5 better than DeepSeek V4 Pro 0424 for agents and tool use?
- Claude Sonnet 5.5 scores higher in agentic tasks (70 vs 64 on the category index, where 50 is average).
- Which is better for long documents?
- Claude Sonnet 5.5 scores higher on long-context benchmarks (67 vs 62). Both accept 1,000,000 tokens of context.
- Which has lower latency?
- DeepSeek V4 Pro 0424 (max) answers first: 49.0 s against 381.5 s to the first answer token, as Artificial Analysis measures it. Reasoning effort drives most of this: a higher setting thinks longer before it answers.
- Which is faster, Claude Sonnet 5.5 or DeepSeek V4 Pro 0424?
- Claude Sonnet 5.5 streams faster: 134 against 92 output tokens per second.
- Do independent benchmarks agree on Claude Sonnet 5.5 vs DeepSeek V4 Pro 0424?
- Across the 13 benchmarks both have, Claude Sonnet 5.5 leads 13, DeepSeek V4 Pro 0424 leads 0. Every overall index that rates both puts Claude Sonnet 5.5 ahead.
- When was Claude Sonnet 5.5 released?
- Claude Sonnet 5.5 was released on 28 Sept 2026.