{"source":"Quelle IA, le classement des modèles de langage","page":"https://quelleia.com/benchmarks/osworld_2/","edition":"2026-W40","date":"2026-09-28","echelle":"score_publie est le score de la source, dans son unité. score_0_1 est ce score ramené entre 0 et 1 pour le calcul. niveau_suggere est le score IA que cette seule mesure indique : 100 pour le meilleur modèle au lancement du site, 50 pour GPT-4o de novembre 2024. difficulte est le score IA d'un modèle qui obtient la moitié (score_0_1 = 0,5), pente la raideur de la courbe.","licence":"Réutilisation libre en citant « Quelle IA, quelleia.com » et la date de l'édition. Les mesures renvoient à leurs sources d'origine.","benchmark":{"id":"osworld_2","nom":"OSWorld 2.0","tache":"agents","etat":"actif","echelle":"unite","modeles":9,"poids_general":1.67,"difficulte":101,"pente":0.1457,"en_une_phrase":"108 longues tâches sur un vrai ordinateur, dans des logiciels professionnels, qui prennent environ une heure et demie à une personne compétente.","notation":"Part de tâches entièrement réussies avec un budget de 500 étapes ; un score partiel compte les points de contrôle validés.","limites":"Scores bas en tout ou rien, 31 % au mieux. Les résultats dépendent du réglage des outils et du niveau de réflexion choisi.","mainteneur":"XLANG Lab (université de Hong Kong)","url":"https://osworld-v2.xlang.ai/","verifie":true},"tache":{"id":"agents","nom":"Agents","poids_general":15,"benchmarks_actifs":9},"modeles_mesures":9,"mesures":16,"classement":[{"rang":1,"modele":"claude-opus-5","nom":"Claude Opus 5","editeur":"Anthropic","page":"https://quelleia.com/modeles/claude-opus-5/","score_publie":0.3143,"score_0_1":0.3143,"niveau_suggere":95.6,"reflexion":"maximale","harnais":null,"configurations_mesurees":5,"date":null,"source":"https://osworld-v2.xlang.ai/"},{"rang":2,"modele":"gpt-5-6-sol","nom":"GPT-5.6 Sol","editeur":"OpenAI","page":"https://quelleia.com/modeles/gpt-5-6-sol/","score_publie":0.2734,"score_0_1":0.2734,"niveau_suggere":94.3,"reflexion":"maximale","harnais":null,"configurations_mesurees":1,"date":null,"source":"https://osworld-v2.xlang.ai/"},{"rang":3,"modele":"claude-opus-4-8","nom":"Claude Opus 4.8","editeur":"Anthropic","page":"https://quelleia.com/modeles/claude-opus-4-8/","score_publie":0.206,"score_0_1":0.206,"niveau_suggere":91.7,"reflexion":"maximale","harnais":null,"configurations_mesurees":2,"date":null,"source":"https://osworld-v2.xlang.ai/"},{"rang":4,"modele":"claude-opus-4-7","nom":"Claude Opus 4.7","editeur":"Anthropic","page":"https://quelleia.com/modeles/claude-opus-4-7/","score_publie":0.182,"score_0_1":0.182,"niveau_suggere":90.7,"reflexion":"maximale","harnais":null,"configurations_mesurees":2,"date":null,"source":"https://osworld-v2.xlang.ai/"},{"rang":5,"modele":"gpt-5-5","nom":"GPT-5.5","editeur":"OpenAI","page":"https://quelleia.com/modeles/gpt-5-5/","score_publie":0.13,"score_0_1":0.13,"niveau_suggere":87.9,"reflexion":"maximale","harnais":null,"configurations_mesurees":1,"date":null,"source":"https://osworld-v2.xlang.ai/"},{"rang":6,"modele":"claude-sonnet-4-6","nom":"Claude Sonnet 4.6","editeur":"Anthropic","page":"https://quelleia.com/modeles/claude-sonnet-4-6/","score_publie":0.093,"score_0_1":0.093,"niveau_suggere":85.3,"reflexion":"moyenne","harnais":null,"configurations_mesurees":2,"date":null,"source":"https://osworld-v2.xlang.ai/"},{"rang":7,"modele":"kimi-k2-6","nom":"Kimi K2.6","editeur":"Moonshot","page":"https://quelleia.com/modeles/kimi-k2-6/","score_publie":0.046,"score_0_1":0.046,"niveau_suggere":80.2,"reflexion":"inconnue","harnais":null,"configurations_mesurees":1,"date":null,"source":"https://osworld-v2.xlang.ai/"},{"rang":7,"modele":"minimax-m3","nom":"MiniMax-M3","editeur":"MiniMax","page":"https://quelleia.com/modeles/minimax-m3/","score_publie":0.046,"score_0_1":0.046,"niveau_suggere":80.2,"reflexion":"inconnue","harnais":null,"configurations_mesurees":1,"date":null,"source":"https://osworld-v2.xlang.ai/"},{"rang":9,"modele":"qwen3-7-plus","nom":"Qwen3.7-Plus","editeur":"Alibaba","page":"https://quelleia.com/modeles/qwen3-7-plus/","score_publie":0.0279999999999999,"score_0_1":0.028,"niveau_suggere":76.6,"reflexion":"inconnue","harnais":null,"configurations_mesurees":1,"date":null,"source":"https://osworld-v2.xlang.ai/"}]}