Title: Code and software agents %E2%80%94 Benchmarks | LLM EU
Canonical: https://llmeu.com/fr/benchmarks/code/
Markdown mirror: https://llmeu.com/fr/benchmarks/code.md
Machine-readable: https://llmeu.com/llms.txt (index) - https://llmeu.com/llms-full.txt (full catalogue) - https://llmeu.com/llms-catalogue.json

# Code and software agents

Scores de codage publi%C3%A9s : SWE-bench Verified et SWE-bench Multilingual pour le travail de d%C3%A9p%C3%B4t agentique, LiveCodeBench pour la g%C3%A9n%C3%A9ration de code de style comp%C3%A9titif, et Terminal Bench pour les agents en ligne de commande.

Every score in this set is republished from the model publisher%E2%80%99s own model card or vendor report and names that source per entry. LLM EU has run no evaluation, has not reproduced any score, and has not verified the harness settings. Vendor-reported numbers are self-reported and are labelled as such on the model cards. SWE-bench results depend heavily on the agent scaffold; the NVIDIA entry names its scaffold explicitly, and scores from different scaffolds must not be compared directly.