Codeforces

We haven't written a description for this benchmark yet; see the card notes below for how each maker ran it.

How each maker reported it

DeepSeek-V4.1-Flash Table 3, p.33
All models at max reasoning effort. Code-agent benchmarks use DeepSeek Harness Minimal mode with 1M context; DeepSWE uses mini-SWE; visual agent tasks use the Claude Code harness with 512K context. HLE for some models is the text-only subset. The Hugging Face card lists NL2Repo-Bench as 64.0; the report and change log say 65.4.
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.

Only each model's own maker's number is shown; scores a maker reports for rival models are left out.