- 1Gemma 4 31B · Google85.2%
- 2Llama 4 Maverick · Meta80.5%
- 3Llama 4 Scout · Meta74.3%
How each maker reported it
- Gemma 4 31B Benchmark results
Instruction-tuned models, as published in Google's model card. OmniDocBench is edit distance (lower is better). Audio benchmarks are omitted because 31B has no audio input.
- Llama 4 Maverick Benchmarks: Instruction tuned models
Meta's own 0-shot results on bf16 weights. Llama 3.x models have no image support and a 128K context, so those cells are not reported.
- Llama 4 Scout Benchmarks: Instruction tuned models
Meta's own 0-shot results on bf16 weights. Llama 3.x models have no image support and a 128K context, so those cells are not reported.
Only each model's own maker's number is shown; scores a maker reports for rival models are left out.