BlogLLM Benchmark Scores as Published, Side by Side
MaXIFE: Results as Published
Every result the catalog holds for MaXIFE: 12 models from 5 developers, read from 2 pages. Each number is as printed by the company named beside it; nothing is ranked or rescored. Models are listed by developer, newest first.
Area MultilingualMaintainer Yile Liu and co-authors (Waseda University and OPPO AI Center)Introduced 2025Data version 2026103001, as of 2026-10-30
MaXIFE tests whether a model follows verifiable instructions in 23 languages. Each example joins one of 795 basic questions with one to three instructions, 1,667 instructions in a language, and the non-English text is a checked translation of the English set, giving 18,285 examples in all. Instructions about format, repetition, keywords, length, and similar constraints are scored by a program, either pass-or-fail or on a tiered point scale. Instructions about style, tone, or switching language are scored by a judge model.
MaXIFE
| Model | Value | Printed by | Setup as printed |
|---|---|---|---|
| GPT-5.2OpenAI, released 2025-12-11 | 88.4% | Alibaba, Qwen3.5-397B-A17B model cardchecked 2026-10-15 | - |
| GPT-5 miniOpenAI, released 2025-08-07 | 85.3% | Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15 | - |
| gpt-oss-120bOpenAI, released 2025-08-05 | 83.7% | Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15 | - |
| Claude Opus 4.5Anthropic, released 2025-11-24 | 79.2% | Alibaba, Qwen3.5-397B-A17B model cardchecked 2026-10-15 | - |
| Gemini 3 ProGoogle, released 2025-11-18 | 87.5% | Alibaba, Qwen3.5-397B-A17B model cardchecked 2026-10-15 | - |
| Qwen3.5-122B-A10BAlibaba, released 2026-02-24 | 87.9% | Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15 | - |
| Qwen3.5-27BAlibaba, released 2026-02-24 | 88% | Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15 | - |
| Qwen3.5-35B-A3BAlibaba, released 2026-02-24 | 86.6% | Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15 | - |
| Qwen3.5-397B-A17BAlibaba, released 2026-02-16 | 88.2% | Alibaba, Qwen3.5-397B-A17B model cardchecked 2026-10-15 | - |
| Qwen3-235B-A22BAlibaba, released 2025-04-29 | 83.2% | Alibaba, Qwen3.5-122B-A10B model cardchecked 2026-10-15 | - |
| Qwen3-Max-ThinkingAlibaba | 84% | Alibaba, Qwen3.5-397B-A17B model cardchecked 2026-10-15 | - |
| Kimi K2.5Moonshot AI, released 2026-01-27 | 72.8% | Alibaba, Qwen3.5-397B-A17B model cardchecked 2026-10-15 | - |
Every number here is read from the same catalog file as the benchmark table, data version 2026103001, and is the number the table shows for that model and test. Where one page printed a model under several setups, each value is listed. The catalog is public in armanobosyan/sugra-llm-benchmarks; a wrong number can be reported from the table, with Report an error under the value, or as an issue in that repository.