How many models are on the cwe-bench leaderboard?

4 entries as of November 8, 2026. cwe-bench is classified as Agents & terminal.

Who is #1 on cwe-bench?

deepseek-ai/DeepSeek-V4.1-Flash-1017 is ranked first on the cwe-bench leaderboard with 57, 9.6% ahead of tencent/Hy4-preview. Read live from the Hugging Face API on November 8, 2026.

Field: Agents & terminal. 4 leaderboard entries. Figures read from the Hugging Face API on November 8, 2026 and refreshed daily.

Top 4

#	Model	Organization	Score
1	deepseek-ai/DeepSeek-V4.1-Flash-1017	deepseek-ai	57
2	tencent/Hy4-preview	tencent	52
3	mistralai/Mistral-Large-4-1T-A52B	mistralai	46.3
4	zai-org/GLM-5.3	zai-org	44

About this board

cwe-bench is a dataset that Hugging Face marks as an official benchmark. Its leaderboard is built automatically from the .eval_results files that model repositories publish, so entries are mostly self-reported by model authors and evaluation settings differ between submissions. A verified tag appears only where Hugging Face marks an entry as verified.

Original board: huggingface.co/datasets/collinear-ai/cwe-bench

Other official benchmarks

hle gpqa typed-decisions physical-engineerings-last-exam SWE-bench_Verified decision-index-leaderboard open-asr-leaderboard olmOCR-bench deep-swe safeagents-judge-bench MMLU-Pro apex-agents agents-last-exam terminal-bench

Part of the Hugging Face official benchmarks map by QuantID. Unofficial. Not affiliated with or endorsed by Hugging Face.