Claude Opus 4.7 vs DeepSeek V4 Pro 0813: Benchmark Comparison
Claude Opus 4.7 has the higher score on 9 of 16 shared benchmarks; DeepSeek V4 Pro 0813 leads on 6.
The largest observed score gap is 14.40 pts on SWE-bench , where DeepSeek V4 Pro 0813 leads.
Reported ±1 standard-error ranges overlap on 7 of 16 shared benchmarks with comparable uncertainty data. This is not a pairwise statistical significance test.
Shared benchmark results
Scores and reported standard errors come from the latest visible Vals benchmark version available for both models.
| Benchmark | Claude Opus 4.7 | DeepSeek V4 Pro 0813 | Gap | Reported uncertainty |
|---|---|---|---|---|
| Harvey's Legal Agent Benchmark | 6.67% ±2.00 | 7.50% ±2.15 | 0.83 pts | Reported ±1 SE ranges overlap |
| Legal Research Bench | 38.46% ±3.38 | 40.87% ±3.42 | 2.40 pts | Reported ±1 SE ranges overlap |
| LegalBench | 85.25% ±0.45 | 82.36% ±0.44 | 2.89 pts | Reported ±1 SE ranges do not overlap |
| EMB | 63.73% ±3.11 | 52.80% ±3.06 | 10.93 pts | Reported ±1 SE ranges do not overlap |
| Finance Agent (v2) | 51.51% ±0.49 | 50.39% ±0.25 | 1.12 pts | Reported ±1 SE ranges do not overlap |
| Tax Agent Bench | 63.86% ±3.14 | 58.66% ±3.14 | 5.20 pts | Reported ±1 SE ranges overlap |
| TaxEval v2 | 75.27% ±0.84 | 73.06% ±0.87 | 2.21 pts | Reported ±1 SE ranges do not overlap |
| MedCode | 54.86% ±2.21 | 42.47% ±2.16 | 12.39 pts | Reported ±1 SE ranges do not overlap |
| MedScribe | 82.95% ±1.98 | 80.17% ±2.00 | 2.78 pts | Reported ±1 SE ranges overlap |
| GPQA Diamond | 90.15% ±1.54 | 92.42% ±2.02 | 2.27 pts | Reported ±1 SE ranges overlap |
| MMLU Pro | 89.87% ±0.30 | 86.97% ±0.34 | 2.90 pts | Reported ±1 SE ranges do not overlap |
| Code Migration | 43.88% ±4.22 | 41.54% ±4.30 | 2.34 pts | Reported ±1 SE ranges overlap |
| LiveCodeBench | 85.07% ±1.01 | 87.53% ±0.96 | 2.45 pts | Reported ±1 SE ranges do not overlap |
| ProgramBench | 0.00% ±0.00 | 0.00% ±0.00 | 0.00 pts | Reported ±1 SE ranges overlap |
| SWE-bench | 82.00% ±1.72 | 96.40% ±0.83 | 14.40 pts | Reported ±1 SE ranges do not overlap |
| Vibe Code Bench v1.1 | 71.00% ±4.51 | 82.30% ±3.15 | 11.29 pts | Reported ±1 SE ranges do not overlap |
Performance by category
| Category | Claude Opus 4.7 average | DeepSeek V4 Pro 0813 average |
|---|---|---|
| Legal | 43.46% | 43.58% |
| Finance | 63.59% | 58.73% |
| Healthcare | 68.91% | 61.32% |
| Academic | 90.01% | 89.70% |
| Coding | 56.39% | 61.55% |
Cost and latency
Cost per test appears only where the benchmark reports it for both models. Latency is the measured completion time for that benchmark.
| Benchmark | Claude Opus 4.7 cost | DeepSeek V4 Pro 0813 cost | Claude Opus 4.7 latency | DeepSeek V4 Pro 0813 latency |
|---|---|---|---|---|
| Harvey's Legal Agent Benchmark | $10.81 | $0.17 | 24m44s | 21m53s |
| Legal Research Bench | $3.06 | $1.09 | 8m32s | 42m07s |
| LegalBench | N/A | N/A | 26.62s | 20.05s |
| EMB | $12.76 | $1.08 | 38m36s | 37m27s |
| Finance Agent (v2) | $4.03 | $0.88 | 6m00s | 16m15s |
| Tax Agent Bench | $1.27 | $0.73 | 4m39s | 26m37s |
| TaxEval v2 | N/A | N/A | 62.07s | 3m46s |
| MedCode | N/A | N/A | 54.25s | 3m01s |
| MedScribe | N/A | N/A | 67.50s | 2m36s |
| GPQA Diamond | N/A | N/A | 3m23s | 4m39s |
| MMLU Pro | N/A | N/A | 3m08s | 59.62s |
| Code Migration | $36.39 | $18.58 | 1h12m | 3h32m |
| LiveCodeBench | N/A | N/A | 7m42s | 4m33s |
| ProgramBench | N/A | N/A | 54m07s | 1h23m |
| SWE-bench | $2.42 | $0.10 | 7m22s | 4m00s |
| Vibe Code Bench v1.1 | $21.41 | $0.36 | 35m42s | 1h06m |
Results available only for Claude Opus 4.7
- Vals RSI Index
- MortgageTax
- MMMU Pro
- SAGE
Results available only for DeepSeek V4 Pro 0813
- Vals Index
- ProofBench v1.1
- Terminal-Bench Science
- IOI
- SkillsBench
- Terminal-Bench 4.0
- Vibe Code Bench 1-100
- CyberBench v1.1