multimodal benchmark
MM IF-Eval
A challenging multimodal instruction-following benchmark that includes both compose-level constraints for output responses and perception-level constraints tied to input images, with comprehensive evaluation pipeline.
Updated Aug 7, 2026
Published models1
Registry coverage1
MetricScore
EvidenceB
MM IF-Eval leaderboard
Sorted by the source-provided rank. Higher score is better according to the registry.
| 1 | MA | 52.7% | 100.0% | 1 | C |
MM IF-Eval highlights
The top published results on this benchmark's own scale.
What is MM IF-Eval?
Definition and scoring fields from the benchmark registry.
A challenging multimodal instruction-following benchmark that includes both compose-level constraints for output responses and perception-level constraints tied to input images, with comprehensive evaluation pipeline.
Scores are shown in ratio. The current registry marks this benchmark as not independently verified with evidence level B.
- Family
- MM IF-Eval
- Modality
- multimodal
- Primary category
- multimodal
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- mm-if-eval|llm-stats-current
Source-native results are preserved. Eligibility for the overall LLMBoard score is a separate policy decision.
FAQ
Common questions about MM IF-Eval.
Which model scores highest on MM IF-Eval?
Pixtral-12B is currently ranked first with 52.7%.
What does MM IF-Eval measure?
A challenging multimodal instruction-following benchmark that includes both compose-level constraints for output responses and perception-level constraints tied to input images, with comprehensive evaluation pipeline.
Is a higher score better?
Yes. Higher values rank better for this benchmark.
How many models are compared?
1 unique published model results are currently shown.
Does this benchmark affect the overall score?
This benchmark is preserved as source-native evidence but is not eligible for the current overall score.