DeepSeek V4 Pro (Reasoning, Max Effort): benchmark results
DeepSeek V4 Pro reasoning max-effort evaluation mode. Provider: DeepSeek. Released 2026-04-23. Access: Open.
Unified ELO 1664 ± 1, rank #201 of 1761 rated models, from 78 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Qwen3.7 Launch - CritPT | 12.9 | Score (%) | 100 |
| Qwen3.7 Launch - LiveCodeBench | 93.5 | Score (%) | 100 |
| Qwen3.7 Launch - Vitabench | 51.9 | Score (%) | 100 |
| AA Omniscience - Software Engineering (SWE) - PHP | 64 | Accuracy (%) | 97.6 |
| AA Omniscience - Software Engineering (SWE) - C | 74 | Accuracy (%) | 97.3 |
| AA Omniscience - Software Engineering (SWE) - Java | 39.39 | Accuracy (%) | 97.3 |
| AA Omniscience - Software Engineering (SWE) - JavaScript | 57.27 | Accuracy (%) | 96.9 |
| AA TAU-2 Bench | 96.2 | Accuracy (%) | 96.8 |
| AA Omniscience - Software Engineering (SWE) - TypeScript | 57.78 | Accuracy (%) | 96.6 |
| AA Omniscience - Software Engineering (SWE) - Dart | 42 | Accuracy (%) | 96.4 |
| AA Omniscience - Software Engineering (SWE) - Rust | 75 | Accuracy (%) | 95.9 |
| AA Omniscience - Software Engineering (SWE) - Go | 46 | Accuracy (%) | 95.6 |
Interactive version: theaggregate.ai/model?slug=deepseek-v4-pro-reasoning-max-effort · How It Works · Data refreshed daily, snapshot 2026-09-05.