An instruction-following benchmark with 58 verifiable single-turn tasks that test generalization to novel and complex constraints.
Browse the latest scores, model modes, release dates, and parameter sizes for IF Bench.
Data sourced primarily from official releases (GitHub, Hugging Face, papers), then benchmark leaderboards, then third-party evaluators. Learn about our data methodology
| Rank | Model | License | |||
|---|---|---|---|---|---|
 Claude Opus 4.6 Extended Thinking | 94.00 | 2026-02-05 | Unknown | Closed | |
 Qwen3.7-Max-Preview Thinking Level · Max | 79.10 | 2026-05-20 | 1000B | Closed | |
Qwen3.5-397B-A17B Thinking Enabled | 76.50 | 2026-02-16 | 39.7B | Free Commercial | |
4 | Qwen3.5-27B Thinking Enabled | 76.50 | 2026-02-25 | 27B | Free Commercial |
5 |  MiniMax-M2.7 Thinking EnabledTools | 76.00 | 2026-03-18 | 229B | Non-Commercial |
6 | Qwen 3.6 Plus Preview Thinking Enabled | 74.20 | 2026-03-31 | Unknown | Closed |
7 | Qwen3.6-Max-Preview Thinking Level · Max | 74.20 | 2026-04-20 | 1000B | Closed |
8 |  GPT-5 Thinking Level · High | 73.10 | 2025-08-07 | Unknown | Closed |
9 | MiniMax M2 Thinking Enabled | 72.30 | 2025-10-27 | 230B | Free Commercial |
10 |  GLM-5 Thinking EnabledTools | 72.00 | 2026-02-11 | 744B | Free Commercial |
11 | Qwen3-Max-Thinking Thinking EnabledTools | 70.90 | 2026-01-26 | 1000B | Closed |
12 |  Gemini 3.0 Pro (Preview 11-2025) Thinking Enabled | 70.00 | 2025-11-18 | Unknown | Closed |
13 | Gemini 3.0 Pro (Preview 11-2025) Thinking Level · HighTools | 70.00 | 2025-11-18 | Unknown | Closed |
14 | M2.1 Thinking EnabledTools | 70.00 | 2025-12-23 | 230B | Free Commercial |
15 | MiniMax M2.5 Thinking EnabledTools | 70.00 | 2026-02-12 | 229B | Free Commercial |
16 | GPT OSS 120B Standard Mode | 69.00 | 2025-08-06 | 11.7B | Free Commercial |
17 |  Kimi K2 Thinking Thinking Enabled | 68.00 | 2025-11-06 | 1040B | Free Commercial |
18 | GLM-4.7-Flash Thinking EnabledTools | 68.00 | 2026-01-19 | 31B | Free Commercial |
19 | GPT OSS 20B Thinking Level · High | 65.10 | 2025-08-06 | 21B | Free Commercial |
20 | Opus 4.5 Extended ThinkingTools | 58.00 | 2025-11-25 | Unknown | Closed |
21 | Claude Sonnet 4.5 Thinking EnabledTools | 57.30 | 2025-09-30 | Unknown | Closed |
22 | Claude Sonnet 4 Thinking EnabledTools | 55.00 | 2025-05-23 | Unknown | Closed |
23 | Opus 4.1 Extended ThinkingTools | 55.00 | 2025-08-06 | Unknown | Closed |
24 | GLM-4.7-Flash Standard ModeTools | 55.00 | 2026-01-19 | 31B | Free Commercial |
25 | Haiku 4.5 Extended Thinking | 54.30 | 2025-10-15 | Unknown | Closed |
26 |  DeepSeek V3.2-Exp Thinking Enabled | 54.10 | 2025-09-29 | 671B | Free Commercial |
27 |  Grok 4.1 Fast Thinking EnabledTools | 53.00 | 2025-11-19 | Unknown | Closed |
28 | Gemini 2.5-Pro Thinking EnabledTools | 49.00 | 2025-06-05 | Unknown | Closed |
29 | GLM-4.6 Thinking Enabled | 43.00 | 2025-09-30 | 355B | Free Commercial |
