DeepSeek
DeepSeek V4 Pro (Non-reasoning) is a DeepSeek model profile, from a family known for strong reasoning, coding, and cost-conscious API models such as R1 and V-series releases. This page separates the public release context from measured benchmark performance so you can inspect its quality and efficiency directly.
DeepSeek release notesRank #137 across 526
Rank #65 across 436
Rank #146 across 357
Percentile score by analysis domain.
* Cost is inverted: lower input, output, and blended prices rank higher.
Higher bars mean stronger relative placement.
| Metric | Domain | Value | Rank |
|---|---|---|---|
| Artificial Analysis Intelligence Index | overall | 31.2 | #137 |
| Artificial Analysis Coding Index | coding | 38.4 | #65 |
| GPQA | reasoning | 71.7% | #205 |
| Humanity's Last Exam | reasoning | 7.7% | #209 |
| SciCode |
| coding, reasoning |
| 42.4% |
| #75 |
| Output Speed | speed | 67 tok/s | #190 |
| Time to First Token | speed | 1.19s | #157 |
| Blended Price | cost | $0.544/M | #146 |
| Input Price | cost | $0.435/M | #183 |
| Output Price | cost | $0.870/M | #123 |
| Value Index | cost, overall | 57.4 | #111 |