Claude Opus 4.7 vs DeepSeek V4 Pro
| Spec | Claude Opus 4.7 | DeepSeek V4 Pro |
|---|---|---|
| Respan Index | 50.0 (#43) | 49.2 (#45) |
| Blended price per 1M | $10 | $0.585 |
| Context window | 1M | 1M |
| Input / output per 1M | $5 / $25 | $0.28 / $1.5 |
| Max output | 128K | 384K |
| Input | text, vision | text |
| Released | 2026-04-16 | 2026-08-13 |
Respan Index by category
Each category is scored 0 to 100, and how far ahead the winner is. Leads under a point are close to even.
| Category | Claude Opus 4.7 | DeepSeek V4 Pro |
|---|---|---|
| Coding | 32.2 Claude Opus 4.7 | 29.8 DeepSeek V4 Pro |
| Agents | 61.1 Claude Opus 4.7 | 59.7 DeepSeek V4 Pro |
| Reasoning | 53.6 Claude Opus 4.7 | 52.8 DeepSeek V4 Pro |
| Knowledge & language | 49.9 Claude Opus 4.7 | 52.6 DeepSeek V4 Pro |
| Human preference | 55.4 Claude Opus 4.7 | 55.7 DeepSeek V4 Pro |
Benchmarks both report
DeepSeek V4 Pro leads on 13 of the 25 benchmarks both report. Higher is better. Each score links to where it was published.
| Benchmark | Claude Opus 4.7 | DeepSeek V4 Pro |
|---|---|---|
| LMArena Elo | 1501 | 1464 |
| LMArena WebDev (LMArena) | 1557 | 1581 |
| MirrorCode (Epoch AI) | 43.3% | 33.3% |
| Chess Puzzles (Epoch AI) | 41% | 46% |
| APEX-Agents (Mercor) | 49.2% | 47.3% |
| FrontierMath Tier 4 v2 (Epoch AI) | 31.7% | 26.8% |
| FrontierMath Tiers 1-3 v2 (Epoch AI) | 70.2% | 64.6% |
| GPQA Diamond | 94.2% | 91.7% |
| LiveBench Agentic Coding (LiveBench) | 50.7% | 54.9% |
| LiveBench Coding (LiveBench) | 82.1% | 77.2% |
| LiveBench Data Analysis (LiveBench) | 78.3% | 79.2% |
| LiveBench Instruction Following (LiveBench) | 66.7% | 67.7% |
| LiveBench Language (LiveBench) | 77.9% | 82.1% |
| LiveBench Mathematics (LiveBench) | 92.9% | 95.1% |
| LiveBench Reasoning (LiveBench) | 87.2% | 85.8% |
| Mystery Game Puzzles (Epoch AI) | 28% | 43% |
| OTIS Mock AIME 2024-2025 (Epoch AI) | 97.8% | 98.6% |
| SimpleQA Verified | 51.7% | 52.9% |
| SWE-bench Verified (Epoch AI) | 83.5% | 77.6% |
| Vending-Bench 2 (Andon Labs) | 10936.76 | 3284.52 |
| WeirdML (Håvard Tveit Ihle) | 76.4% | 66.2% |
| AIME 2026 | 95.8% | 96.7% |
| Humanity's Last Exam (with tools) | 54.7% | 60% |
| Terminal-Bench 2.1 | 66.1% | 87.9% |
| SimpleBench (SimpleBench) | 61.7% | 50.9% |
FAQ
Is Claude Opus 4.7 better than DeepSeek V4 Pro?
Claude Opus 4.7 scores 50.0 on the Respan Index (#43), ahead of DeepSeek V4 Pro at 49.2 (#45). DeepSeek V4 Pro leads on 13 of the 25 benchmarks both report.
Which is cheaper, Claude Opus 4.7 or DeepSeek V4 Pro?
Claude Opus 4.7 lists at $5 input and $25 output per 1M tokens; DeepSeek V4 Pro at $0.28 and $1.5. DeepSeek V4 Pro is 17x cheaper per token at list price.
Which has the larger context window?
Both accept 1M tokens of context.
Compare other models
- vsClaude Opus 4.7 vs Claude Opus 4.6
- vsClaude Opus 4.8 vs Claude Opus 4.7
- vsClaude Sonnet 5.5 vs DeepSeek V4 Pro
- vsGemini 3.8 Flash vs DeepSeek V4 Pro
- vsClaude Opus 4.7 vs GPT-5.4 Pro
- vsClaude Opus 4.7 vs GPT-5.4
- vsClaude Opus 4.7 vs MiMo-V2.6-Flash
- vsDeepSeek V4 Pro vs GPT-5.4
- vsDeepSeek V4 Pro vs GPT-6 Luna
- vsDeepSeek V4 Pro vs GPT-5.6 Luna