Complex Instruction Following on Internal Evaluation
0Tie ScoreDeepSeek-V4-Pro
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| DeepSeek-V4-Pro2026.04 | 0 | 23 | — | 46.9 | — | 0 | |
| Claude-Opus-4.52026.04 | 0 | — | 26 | — | 53.1 | 0 |
Loading the SOTA2 catalog…
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| DeepSeek-V4-Pro2026.04 | 0 | 23 | — | 46.9 | — | 0 | |
| Claude-Opus-4.52026.04 | 0 | — | 26 | — | 53.1 | 0 |