Benchmark index
Every benchmark we run. A new benchmark opens roughly every two weeks. Live benchmarks accumulate judgements on a fixed prompt set; each release is versioned and permanently citable.
Live
Live benchmarks update continuously; frozen rounds remain available as citable snapshots from the benchmark page.
Coming soon
Text renderingComing soon
Text-to-Image
Text accuracy, legibility, layout
Variance (Mode Collapse)Coming soon
Text-to-Image
Output diversity across seeds
Image editingComing soon
Image-to-image
Instruction adherence, edit quality
Facial preservationComing soon
Image-to-image
Identity retention
General videoComing soon
Text-to-Video
Preference, temporal coherence
SVG generationComing soon
SVG
Structure, readability, logo fidelity
Animated SVGComing soon
SVG
Motion fidelity
General 3DComing soon
3D
Geometry, texture, view consistency
Want a benchmark that is not listed? Suggest a benchmark.
We use crowd intelligence to deliver real, diverse human feedback at scale for model evaluation and post-training.
Rapidata, benchmark.ai © 2026