{"source":"Quelle IA, le classement des modèles de langage","page":"https://quelleia.com/modeles/gpt-5-5/","edition":"2026-W40","date":"2026-09-28","echelle":"Score IA : 100 pour le meilleur modèle au lancement du site, 50 pour GPT-4o de novembre 2024. bas et haut bornent l'intervalle à 90 %.","licence":"Réutilisation libre en citant « Quelle IA, quelleia.com » et la date de l'édition. Les mesures renvoient à leurs sources d'origine.","modele":{"liens":[{"type":"essayer","libelle":"Essayer sur ChatGPT","url":"https://chatgpt.com"},{"type":"openrouter","libelle":"Utiliser par API sur OpenRouter","url":"https://openrouter.ai/openai/gpt-5.5"},{"type":"api","libelle":"Documentation de l'API (OpenAI)","url":"https://platform.openai.com/docs"},{"type":"editeur","libelle":"Site de l'éditeur (OpenAI)","url":"https://openai.com"}],"id":"gpt-5-5","nom":"GPT-5.5","editeur":"OpenAI","pays":"États-Unis","date_sortie":"2026-04-23","nouveau":false,"poids_ouverts":false,"acces":"API access","licence":"proprietary","local":null,"prix_eur_par_million":{"entree":4.3354,"sortie":26.0122},"hebergeable_ue":true,"classe_energie":"E","contexte_jetons":1050000,"entrees":["text","image","pdf"],"niveaux_reflexion":["none","low","medium","high","xhigh"],"connaissances_jusqu_a":"2025-12-01","general":{"score":91.3,"bas":90,"haut":92.5,"mesures":45,"poids_couvert":100,"sources":38,"provisoire":false,"historique":false,"rang":11,"groupe":4},"taches":{"raisonnement":{"score":94.1,"bas":91.6,"haut":96.6,"mesures":8,"benchmarks_tache":9,"couverture":89,"provisoire":false,"rang":8},"code":{"score":90.2,"bas":88.6,"haut":91.8,"mesures":10,"benchmarks_tache":13,"couverture":77,"provisoire":false,"rang":17},"agents":{"score":90.7,"bas":88.1,"haut":93.3,"mesures":7,"benchmarks_tache":9,"couverture":78,"provisoire":false,"rang":6},"usage":{"score":89.9,"bas":81.8,"haut":98.1,"mesures":1,"benchmarks_tache":1,"couverture":100,"provisoire":false,"rang":12},"maths":{"score":91.5,"bas":88.6,"haut":94.5,"mesures":8,"benchmarks_tache":8,"couverture":100,"provisoire":false,"rang":11},"sciences":{"score":94.6,"bas":89.9,"haut":99.4,"mesures":4,"benchmarks_tache":5,"couverture":80,"provisoire":false,"rang":7},"francais":{"score":91.5,"bas":82,"haut":101,"mesures":2,"benchmarks_tache":10,"couverture":50,"provisoire":false,"rang":7},"vision":{"score":92.6,"bas":90.4,"haut":94.7,"mesures":3,"benchmarks_tache":3,"couverture":100,"provisoire":false,"rang":6},"ecriture":{"score":89.9,"bas":81.8,"haut":98.1,"mesures":1,"benchmarks_tache":1,"couverture":100,"provisoire":false,"rang":16},"web":{"score":85.5,"bas":80.5,"haut":90.6,"mesures":1,"benchmarks_tache":1,"couverture":100,"provisoire":false,"rang":40}},"reflexions_mesurees":["elevee","faible","maximale","moyenne","sans"],"harnais_mesures":["Capy","Codex CLI","NexAU-AHE","OpenHands","clnkr","mini-SWE-agent"],"mesures":[{"benchmark":"otis_mock_aime_2024_2025","tache":"maths","etat":"sature","score_publie":1,"score_0_1":1,"reflexion":"maximale","harnais":null,"provenance":"independant","date":"2026-04-24","source":"https://epoch.ai/benchmarks","niveau_suggere":93.8},{"benchmark":"arc_agi","tache":"raisonnement","etat":"sature","score_publie":0.95,"score_0_1":0.95,"reflexion":"maximale","harnais":null,"provenance":"externe","date":null,"source":"https://arcprize.org/leaderboard","niveau_suggere":91.3},{"benchmark":"arc_agi","tache":"raisonnement","etat":"sature","score_publie":0.95,"score_0_1":0.95,"reflexion":"maximale","harnais":null,"provenance":"externe","date":null,"source":"https://arcprize.org/leaderboard","niveau_suggere":91.3},{"benchmark":"arc_agi","tache":"raisonnement","etat":"sature","score_publie":0.945,"score_0_1":0.945,"reflexion":"elevee","harnais":null,"provenance":"externe","date":null,"source":"https://arcprize.org/leaderboard","niveau_suggere":90.7},{"benchmark":"dtbench","tache":"raisonnement","etat":"sature","score_publie":0.96,"score_0_1":0.9333,"reflexion":"maximale","harnais":null,"provenance":"externe","date":null,"source":"https://conceptualreasoning.ai/dtbench","niveau_suggere":99.5},{"benchmark":"arc_agi","tache":"raisonnement","etat":"sature","score_publie":0.9217,"score_0_1":0.9217,"reflexion":"moyenne","harnais":null,"provenance":"externe","date":null,"source":"https://arcprize.org/leaderboard","niveau_suggere":88.5},{"benchmark":"gpqa_diamond","tache":"sciences","etat":"actif","score_publie":0.9400252525252524,"score_0_1":0.92,"reflexion":"maximale","harnais":null,"provenance":"independant","date":"2026-04-24","source":"https://epoch.ai/benchmarks","niveau_suggere":89.5},{"benchmark":"frontiermath","tache":"maths","etat":"actif","score_publie":0.517,"score_0_1":0.907,"reflexion":"maximale","harnais":null,"provenance":"independant","date":"2026-04-23","source":"https://epoch.ai/benchmarks","niveau_suggere":94},{"benchmark":"surface_evolver_bench","tache":"sciences","etat":"sature","score_publie":0.88125,"score_0_1":0.8812,"reflexion":"elevee","harnais":null,"provenance":"externe","date":null,"source":"https://yhenon.github.io/surface-evolver-llm-eval/","niveau_suggere":98},{"benchmark":"gpqa_diamond","tache":"sciences","etat":"actif","score_publie":0.9065656565656566,"score_0_1":0.8754,"reflexion":"faible","harnais":null,"provenance":"independant","date":"2026-05-05","source":"https://epoch.ai/benchmarks","niveau_suggere":83.7},{"benchmark":"frontiermath_tiers_1_3_v2","tache":"maths","etat":"actif","score_publie":0.8526315789473684,"score_0_1":0.8526,"reflexion":"maximale","harnais":null,"provenance":"independant","date":"2026-06-11","source":"https://epoch.ai/benchmarks","niveau_suggere":95.2},{"benchmark":"arc_agi_2","tache":"raisonnement","etat":"sature","score_publie":0.85,"score_0_1":0.85,"reflexion":"maximale","harnais":null,"provenance":"externe","date":null,"source":"https://epoch.ai/benchmarks","niveau_suggere":92.7},{"benchmark":"arc_agi_2","tache":"raisonnement","etat":"sature","score_publie":0.85,"score_0_1":0.85,"reflexion":"maximale","harnais":null,"provenance":"externe","date":null,"source":"https://epoch.ai/benchmarks","niveau_suggere":92.7},{"benchmark":"weirdml","tache":"code","etat":"actif","score_publie":0.8491,"score_0_1":0.8491,"reflexion":"maximale","harnais":null,"provenance":"externe","date":null,"source":"https://htihle.github.io/weirdml.html","niveau_suggere":103.8},{"benchmark":"terminal_bench","tache":"code","etat":"actif","score_publie":0.847191011236,"score_0_1":0.8472,"reflexion":"inconnue","harnais":"NexAU-AHE","provenance":"externe","date":"2026-04-23","source":"https://www.tbench.ai/leaderboard/terminal-bench/2.0","niveau_suggere":91.9},{"benchmark":"otis_mock_aime_2024_2025","tache":"maths","etat":"sature","score_publie":0.8444444444444444,"score_0_1":0.8443,"reflexion":"faible","harnais":null,"provenance":"independant","date":"2026-08-07","source":"https://epoch.ai/benchmarks","niveau_suggere":73.9},{"benchmark":"weirdml","tache":"code","etat":"actif","score_publie":0.8390000000000001,"score_0_1":0.839,"reflexion":"elevee","harnais":null,"provenance":"externe","date":null,"source":"https://htihle.github.io/weirdml.html","niveau_suggere":102.5},{"benchmark":"arc_agi_2","tache":"raisonnement","etat":"sature","score_publie":0.8333,"score_0_1":0.8333,"reflexion":"elevee","harnais":null,"provenance":"externe","date":null,"source":"https://epoch.ai/benchmarks","niveau_suggere":92.2},{"benchmark":"terminal_bench","tache":"code","etat":"actif","score_publie":0.8314606741570001,"score_0_1":0.8315,"reflexion":"inconnue","harnais":"Capy","provenance":"externe","date":"2026-04-23","source":"https://www.tbench.ai/leaderboard/terminal-bench/2.0","niveau_suggere":90.9},{"benchmark":"terminal_bench","tache":"code","etat":"actif","score_publie":0.822471910112,"score_0_1":0.8225,"reflexion":"inconnue","harnais":"Codex CLI","provenance":"externe","date":"2026-04-23","source":"https://www.tbench.ai/leaderboard/terminal-bench/2.0","niveau_suggere":90.4},{"benchmark":"terminal_bench","tache":"code","etat":"actif","score_publie":0.8200000000000001,"score_0_1":0.82,"reflexion":"inconnue","harnais":"Codex CLI","provenance":"externe","date":"2026-04-23","source":"https://www.tbench.ai/leaderboard/terminal-bench/2.0","niveau_suggere":90.3},{"benchmark":"surface_evolver_bench","tache":"sciences","etat":"sature","score_publie":0.8125,"score_0_1":0.8125,"reflexion":"moyenne","harnais":null,"provenance":"externe","date":null,"source":"https://yhenon.github.io/surface-evolver-llm-eval/","niveau_suggere":93.4},{"benchmark":"swe_bench_verified","tache":"code","etat":"actif","score_publie":0.8057851239669421,"score_0_1":0.8058,"reflexion":"maximale","harnais":null,"provenance":"independant","date":"2026-04-24","source":"https://epoch.ai/benchmarks","niveau_suggere":88.8},{"benchmark":"arc_agi","tache":"raisonnement","etat":"sature","score_publie":0.7617,"score_0_1":0.7617,"reflexion":"faible","harnais":null,"provenance":"externe","date":null,"source":"https://arcprize.org/leaderboard","niveau_suggere":80.8},{"benchmark":"frontiermath_tier_4_v2","tache":"maths","etat":"sature","score_publie":0.725,"score_0_1":0.725,"reflexion":"maximale","harnais":null,"provenance":"independant","date":"2026-06-11","source":"https://epoch.ai/benchmarks","niveau_suggere":94.5},{"benchmark":"arc_agi_2","tache":"raisonnement","etat":"sature","score_publie":0.7042,"score_0_1":0.7042,"reflexion":"moyenne","harnais":null,"provenance":"externe","date":null,"source":"https://epoch.ai/benchmarks","niveau_suggere":88.9},{"benchmark":"gpqa_diamond","tache":"sciences","etat":"actif","score_publie":0.7727272727272727,"score_0_1":0.697,"reflexion":"sans","harnais":null,"provenance":"independant","date":"2026-08-07","source":"https://epoch.ai/benchmarks","niveau_suggere":70.6},{"benchmark":"weirdml","tache":"code","etat":"actif","score_publie":0.6715000000000001,"score_0_1":0.6715,"reflexion":"sans","harnais":null,"provenance":"externe","date":null,"source":"https://htihle.github.io/weirdml.html","niveau_suggere":86.7},{"benchmark":"deepswe","tache":"code","etat":"actif","score_publie":0.6703539823008849,"score_0_1":0.6704,"reflexion":"maximale","harnais":"mini-SWE-agent","provenance":"externe","date":null,"source":"https://deepswe.datacurve.ai/","niveau_suggere":91.9},{"benchmark":"terminal_bench","tache":"code","etat":"actif","score_publie":0.660674157303,"score_0_1":0.6607,"reflexion":"inconnue","harnais":"clnkr","provenance":"externe","date":"2026-04-23","source":"https://www.tbench.ai/leaderboard/terminal-bench/2.0","niveau_suggere":83.5},{"benchmark":"deepswe","tache":"code","etat":"actif","score_publie":0.6438053097345133,"score_0_1":0.6438,"reflexion":"elevee","harnais":"mini-SWE-agent","provenance":"externe","date":null,"source":"https://deepswe.datacurve.ai/","niveau_suggere":91},{"benchmark":"lmca","tache":"raisonnement","etat":"actif","score_publie":54.27,"score_0_1":0.6385,"reflexion":"maximale","harnais":null,"provenance":"externe","date":null,"source":"https://conceptualreasoning.ai/lmca","niveau_suggere":91.5},{"benchmark":"arena_maths","tache":"maths","etat":"actif","score_publie":1499.4,"score_0_1":0.6343,"reflexion":"inconnue","harnais":null,"provenance":"independant","date":"2026-09-30","source":"https://arena.ai/leaderboard","niveau_suggere":89.9},{"benchmark":"simpleqa_verified","tache":"donnees","etat":"actif","score_publie":0.63,"score_0_1":0.63,"reflexion":"maximale","harnais":null,"provenance":"independant","date":"2026-08-27","source":"https://epoch.ai/benchmarks","niveau_suggere":96},{"benchmark":"simplebench","tache":"raisonnement","etat":"actif","score_publie":0.6900000000000001,"score_0_1":0.628,"reflexion":"inconnue","harnais":null,"provenance":"externe","date":null,"source":"https://lmcouncil.ai/benchmarks","niveau_suggere":89.5},{"benchmark":"arena_expert","tache":"sciences","etat":"actif","score_publie":1512.4,"score_0_1":0.6266,"reflexion":"elevee","harnais":null,"provenance":"independant","date":"2026-09-30","source":"https://arena.ai/leaderboard","niveau_suggere":87.6},{"benchmark":"arena_maths","tache":"maths","etat":"actif","score_publie":1493.2,"score_0_1":0.626,"reflexion":"elevee","harnais":null,"provenance":"independant","date":"2026-09-30","source":"https://arena.ai/leaderboard","niveau_suggere":88.6},{"benchmark":"arena_expert","tache":"sciences","etat":"actif","score_publie":1511.4,"score_0_1":0.6252,"reflexion":"inconnue","harnais":null,"provenance":"independant","date":"2026-09-30","source":"https://arena.ai/leaderboard","niveau_suggere":87.5},{"benchmark":"comparia","tache":"francais","etat":"actif","score_publie":1110,"score_0_1":0.6199,"reflexion":"inconnue","harnais":null,"provenance":"independant","date":"2026-09-30","source":"https://arene.comparia.beta.gouv.fr/ranking","niveau_suggere":110.6},{"benchmark":"arena_ecriture","tache":"ecriture","etat":"actif","score_publie":1451.4,"score_0_1":0.6137,"reflexion":"elevee","harnais":null,"provenance":"independant","date":"2026-09-30","source":"https://arena.ai/leaderboard","niveau_suggere":88.2},{"benchmark":"arena_texte","tache":"usage","etat":"actif","score_publie":1481.3,"score_0_1":0.6122,"reflexion":"elevee","harnais":null,"provenance":"independant","date":"2026-09-30","source":"https://arena.ai/leaderboard","niveau_suggere":88.2},{"benchmark":"arena_ecriture","tache":"ecriture","etat":"actif","score_publie":1448,"score_0_1":0.609,"reflexion":"inconnue","harnais":null,"provenance":"independant","date":"2026-09-30","source":"https://arena.ai/leaderboard","niveau_suggere":87.4},{"benchmark":"arena_francais","tache":"francais","etat":"actif","score_publie":1501.3,"score_0_1":0.6067,"reflexion":"inconnue","harnais":null,"provenance":"independant","date":"2026-09-30","source":"https://arena.ai/leaderboard","niveau_suggere":87},{"benchmark":"arena_texte","tache":"usage","etat":"actif","score_publie":1476.7,"score_0_1":0.6059,"reflexion":"inconnue","harnais":null,"provenance":"independant","date":"2026-09-30","source":"https://arena.ai/leaderboard","niveau_suggere":87.2},{"benchmark":"arena_webdev","tache":"web","etat":"actif","score_publie":1512.1,"score_0_1":0.6027,"reflexion":"maximale","harnais":null,"provenance":"independant","date":"2026-09-30","source":"https://arena.ai/leaderboard","niveau_suggere":84.9},{"benchmark":"arena_code","tache":"code","etat":"actif","score_publie":1518.6,"score_0_1":0.6013,"reflexion":"elevee","harnais":null,"provenance":"independant","date":"2026-09-30","source":"https://arena.ai/leaderboard","niveau_suggere":85.6},{"benchmark":"arena_francais","tache":"francais","etat":"actif","score_publie":1495.8,"score_0_1":0.5991,"reflexion":"elevee","harnais":null,"provenance":"independant","date":"2026-09-30","source":"https://arena.ai/leaderboard","niveau_suggere":85.9},{"benchmark":"arena_code","tache":"code","etat":"actif","score_publie":1510.5,"score_0_1":0.5901,"reflexion":"inconnue","harnais":null,"provenance":"independant","date":"2026-09-30","source":"https://arena.ai/leaderboard","niveau_suggere":83.9},{"benchmark":"frontiermath_tier_4","tache":"maths","etat":"actif","score_publie":0.354,"score_0_1":0.59,"reflexion":"maximale","harnais":null,"provenance":"independant","date":"2026-04-23","source":"https://epoch.ai/benchmarks","niveau_suggere":92},{"benchmark":"arena_vision","tache":"vision","etat":"actif","score_publie":1286.5,"score_0_1":0.5862,"reflexion":"inconnue","harnais":null,"provenance":"independant","date":"2026-09-28","source":"https://arena.ai/leaderboard","niveau_suggere":89.6},{"benchmark":"arena_vision","tache":"vision","etat":"actif","score_publie":1281,"score_0_1":0.5785,"reflexion":"elevee","harnais":null,"provenance":"independant","date":"2026-09-28","source":"https://arena.ai/leaderboard","niveau_suggere":88.3},{"benchmark":"otis_mock_aime_2024_2025","tache":"maths","etat":"sature","score_publie":0.5777777777777777,"score_0_1":0.5774,"reflexion":"sans","harnais":null,"provenance":"independant","date":"2026-08-07","source":"https://epoch.ai/benchmarks","niveau_suggere":64.5},{"benchmark":"arena_webdev","tache":"web","etat":"actif","score_publie":1487,"score_0_1":0.5677,"reflexion":"elevee","harnais":null,"provenance":"independant","date":"2026-09-30","source":"https://arena.ai/leaderboard","niveau_suggere":83.2},{"benchmark":"scicode","tache":"code","etat":"actif","score_publie":0.561342592592593,"score_0_1":0.5613,"reflexion":"maximale","harnais":null,"provenance":"externe","date":null,"source":"https://artificialanalysis.ai/evaluations/scicode","niveau_suggere":90.5},{"benchmark":"scicode","tache":"code","etat":"actif","score_publie":0.559027777777778,"score_0_1":0.559,"reflexion":"elevee","harnais":null,"provenance":"externe","date":null,"source":"https://artificialanalysis.ai/evaluations/scicode","niveau_suggere":90.2},{"benchmark":"apex_agents","tache":"agents","etat":"actif","score_publie":0.551,"score_0_1":0.551,"reflexion":"inconnue","harnais":null,"provenance":"externe","date":null,"source":"https://epoch.ai/benchmarks","niveau_suggere":90.4},{"benchmark":"deepresearch_bench","tache":"agents","etat":"actif","score_publie":0.54,"score_0_1":0.54,"reflexion":"elevee","harnais":null,"provenance":"externe","date":null,"source":"https://evals.futuresearch.ai/#drb","niveau_suggere":106.8},{"benchmark":"deepswe","tache":"code","etat":"actif","score_publie":0.5398230088495575,"score_0_1":0.5398,"reflexion":"moyenne","harnais":"mini-SWE-agent","provenance":"externe","date":null,"source":"https://deepswe.datacurve.ai/","niveau_suggere":87.7},{"benchmark":"scicode","tache":"code","etat":"actif","score_publie":0.534722222222222,"score_0_1":0.5347,"reflexion":"moyenne","harnais":null,"provenance":"externe","date":null,"source":"https://artificialanalysis.ai/evaluations/scicode","niveau_suggere":87.3},{"benchmark":"gbaeval","tache":"code","etat":"actif","score_publie":0.5322037178151383,"score_0_1":0.5322,"reflexion":"inconnue","harnais":null,"provenance":"externe","date":null,"source":"https://gbaeval.com/model/gpt-5-5","niveau_suggere":90.3},{"benchmark":"arena_webdev","tache":"web","etat":"actif","score_publie":1455.3,"score_0_1":0.5224,"reflexion":"inconnue","harnais":null,"provenance":"independant","date":"2026-09-30","source":"https://arena.ai/leaderboard","niveau_suggere":81},{"benchmark":"vending_bench_2","tache":"agents","etat":"actif","score_publie":7523.843999999999,"score_0_1":0.5178,"reflexion":"inconnue","harnais":null,"provenance":"externe","date":null,"source":"https://epoch.ai/benchmarks","niveau_suggere":89.4},{"benchmark":"scicode","tache":"code","etat":"actif","score_publie":0.516203703703704,"score_0_1":0.5162,"reflexion":"faible","harnais":null,"provenance":"externe","date":null,"source":"https://artificialanalysis.ai/evaluations/scicode","niveau_suggere":85.1},{"benchmark":"chess_puzzles","tache":"raisonnement","etat":"actif","score_publie":0.54,"score_0_1":0.516,"reflexion":"maximale","harnais":null,"provenance":"independant","date":"2026-04-24","source":"https://epoch.ai/benchmarks","niveau_suggere":95.2},{"benchmark":"mystery_game_puzzles","tache":"raisonnement","etat":"actif","score_publie":0.56,"score_0_1":0.5153,"reflexion":"maximale","harnais":null,"provenance":"independant","date":"2026-07-24","source":"https://epoch.ai/benchmarks","niveau_suggere":95.3},{"benchmark":"proofbench","tache":"maths","etat":"sature","score_publie":0.5,"score_0_1":0.5,"reflexion":"maximale","harnais":null,"provenance":"externe","date":null,"source":"https://www.vals.ai/benchmarks/proof_bench","niveau_suggere":87.9},{"benchmark":"deepresearch_bench","tache":"agents","etat":"actif","score_publie":0.496,"score_0_1":0.496,"reflexion":"moyenne","harnais":null,"provenance":"externe","date":null,"source":"https://evals.futuresearch.ai/#drb","niveau_suggere":91.2},{"benchmark":"deepresearch_bench","tache":"agents","etat":"actif","score_publie":0.487,"score_0_1":0.487,"reflexion":"faible","harnais":null,"provenance":"externe","date":null,"source":"https://evals.futuresearch.ai/#drb","niveau_suggere":88},{"benchmark":"exploitbench","tache":"securite","etat":"actif","score_publie":0.474,"score_0_1":0.474,"reflexion":"inconnue","harnais":null,"provenance":"externe","date":null,"source":"https://epoch.ai/benchmarks","niveau_suggere":94.3},{"benchmark":"scicode","tache":"code","etat":"actif","score_publie":0.47337962962963,"score_0_1":0.4734,"reflexion":"sans","harnais":null,"provenance":"externe","date":null,"source":"https://artificialanalysis.ai/evaluations/scicode","niveau_suggere":80.1},{"benchmark":"mystery_game_puzzles","tache":"raisonnement","etat":"actif","score_publie":0.52,"score_0_1":0.4712,"reflexion":"elevee","harnais":null,"provenance":"independant","date":"2026-07-27","source":"https://epoch.ai/benchmarks","niveau_suggere":94.1},{"benchmark":"frontiercode","tache":"code","etat":"actif","score_publie":0.4296,"score_0_1":0.4296,"reflexion":"inconnue","harnais":"Codex CLI","provenance":"externe","date":null,"source":"https://cognition.com/frontiercode","niveau_suggere":91.5},{"benchmark":"exploitbench","tache":"securite","etat":"actif","score_publie":0.418,"score_0_1":0.418,"reflexion":"inconnue","harnais":null,"provenance":"externe","date":null,"source":"https://epoch.ai/benchmarks","niveau_suggere":92.1},{"benchmark":"exploitbench","tache":"securite","etat":"actif","score_publie":0.416,"score_0_1":0.416,"reflexion":"inconnue","harnais":"Codex CLI","provenance":"externe","date":null,"source":"https://epoch.ai/benchmarks","niveau_suggere":92.1},{"benchmark":"gso_bench","tache":"code","etat":"actif","score_publie":0.402,"score_0_1":0.402,"reflexion":"maximale","harnais":"OpenHands","provenance":"externe","date":null,"source":"https://gso-bench.github.io/index.html","niveau_suggere":90},{"benchmark":"blueprint_bench_2","tache":"vision","etat":"actif","score_publie":0.3622559652928415,"score_0_1":0.3623,"reflexion":"inconnue","harnais":null,"provenance":"externe","date":null,"source":"https://andonlabs.com/evals/blueprint-bench-2","niveau_suggere":94.1},{"benchmark":"ebr_bench","tache":"agents","etat":"actif","score_publie":0.3428571428571428,"score_0_1":0.3429,"reflexion":"maximale","harnais":null,"provenance":"independant","date":"2026-07-27","source":"https://epoch.ai/benchmarks","niveau_suggere":91.9},{"benchmark":"arc_agi_2","tache":"raisonnement","etat":"sature","score_publie":0.3333,"score_0_1":0.3333,"reflexion":"faible","harnais":null,"provenance":"externe","date":null,"source":"https://epoch.ai/benchmarks","niveau_suggere":82},{"benchmark":"exploitbench","tache":"securite","etat":"actif","score_publie":0.326,"score_0_1":0.326,"reflexion":"inconnue","harnais":"Codex CLI","provenance":"externe","date":null,"source":"https://epoch.ai/benchmarks","niveau_suggere":88.3},{"benchmark":"posttrainbench","tache":"agents","etat":"actif","score_publie":0.2723,"score_0_1":0.2723,"reflexion":"maximale","harnais":"Codex CLI","provenance":"externe","date":null,"source":"https://epoch.ai/benchmarks","niveau_suggere":87.8},{"benchmark":"critpt","tache":"sciences","etat":"actif","score_publie":0.271428571428571,"score_0_1":0.2714,"reflexion":"maximale","harnais":null,"provenance":"externe","date":null,"source":"https://epoch.ai/benchmarks","niveau_suggere":94.1},{"benchmark":"deepswe","tache":"code","etat":"actif","score_publie":0.2699115044247787,"score_0_1":0.2699,"reflexion":"faible","harnais":"mini-SWE-agent","provenance":"externe","date":null,"source":"https://deepswe.datacurve.ai/","niveau_suggere":78.9},{"benchmark":"gdp_pdf","tache":"donnees","etat":"actif","score_publie":0.26,"score_0_1":0.26,"reflexion":"maximale","harnais":null,"provenance":"externe","date":null,"source":"https://surgehq.ai/benchmarks/gdp-pdf","niveau_suggere":93.8},{"benchmark":"critpt","tache":"sciences","etat":"actif","score_publie":0.254285714285714,"score_0_1":0.2543,"reflexion":"elevee","harnais":null,"provenance":"externe","date":null,"source":"https://epoch.ai/benchmarks","niveau_suggere":93.3},{"benchmark":"cl_bench_life","tache":"donnees","etat":"actif","score_publie":0.222,"score_0_1":0.222,"reflexion":"elevee","harnais":null,"provenance":"externe","date":null,"source":"https://epoch.ai/benchmarks","niveau_suggere":91.7},{"benchmark":"chess_puzzles","tache":"raisonnement","etat":"actif","score_publie":0.26,"score_0_1":0.2214,"reflexion":"faible","harnais":null,"provenance":"independant","date":"2026-08-07","source":"https://epoch.ai/benchmarks","niveau_suggere":81.7},{"benchmark":"forecastbench","tache":"raisonnement","etat":"actif","score_publie":60.6,"score_0_1":0.212,"reflexion":"inconnue","harnais":null,"provenance":"externe","date":null,"source":"https://epoch.ai/benchmarks","niveau_suggere":82.5},{"benchmark":"mystery_game_puzzles","tache":"raisonnement","etat":"actif","score_publie":0.28,"score_0_1":0.2069,"reflexion":"faible","harnais":null,"provenance":"independant","date":"2026-08-28","source":"https://epoch.ai/benchmarks","niveau_suggere":86.4},{"benchmark":"furniture_assembly","tache":"vision","etat":"actif","score_publie":0.4416666666666666,"score_0_1":0.2024,"reflexion":"maximale","harnais":null,"provenance":"independant","date":"2026-09-10","source":"https://epoch.ai/benchmarks","niveau_suggere":92.3},{"benchmark":"critpt","tache":"sciences","etat":"actif","score_publie":0.185714285714286,"score_0_1":0.1857,"reflexion":"moyenne","harnais":null,"provenance":"externe","date":null,"source":"https://epoch.ai/benchmarks","niveau_suggere":89.6},{"benchmark":"osworld_2","tache":"agents","etat":"actif","score_publie":0.13,"score_0_1":0.13,"reflexion":"maximale","harnais":null,"provenance":"externe","date":null,"source":"https://osworld-v2.xlang.ai/","niveau_suggere":87.9},{"benchmark":"mirrorcode","tache":"code","etat":"actif","score_publie":0.1,"score_0_1":0.1,"reflexion":"elevee","harnais":null,"provenance":"independant","date":"2026-08-10","source":"https://epoch.ai/benchmarks","niveau_suggere":86.2},{"benchmark":"mystery_game_puzzles","tache":"raisonnement","etat":"actif","score_publie":0.18,"score_0_1":0.0967,"reflexion":"sans","harnais":null,"provenance":"independant","date":"2026-08-27","source":"https://epoch.ai/benchmarks","niveau_suggere":80.7},{"benchmark":"critpt","tache":"sciences","etat":"actif","score_publie":0.08,"score_0_1":0.08,"reflexion":"faible","harnais":null,"provenance":"externe","date":null,"source":"https://epoch.ai/benchmarks","niveau_suggere":80.8},{"benchmark":"remote_labor_index","tache":"agents","etat":"actif","score_publie":0.0625,"score_0_1":0.0625,"reflexion":"inconnue","harnais":null,"provenance":"externe","date":null,"source":"https://epoch.ai/benchmarks","niveau_suggere":89.6},{"benchmark":"chess_puzzles","tache":"raisonnement","etat":"actif","score_publie":0.1,"score_0_1":0.053,"reflexion":"sans","harnais":null,"provenance":"independant","date":"2026-08-07","source":"https://epoch.ai/benchmarks","niveau_suggere":65.1},{"benchmark":"critpt","tache":"sciences","etat":"actif","score_publie":0.0142857142857143,"score_0_1":0.0143,"reflexion":"sans","harnais":null,"provenance":"externe","date":null,"source":"https://epoch.ai/benchmarks","niveau_suggere":64.5},{"benchmark":"frontiermath_erdos","tache":"maths","etat":"actif","score_publie":0,"score_0_1":0,"reflexion":"maximale","harnais":null,"provenance":"independant","date":"2026-08-28","source":"https://epoch.ai/benchmarks","niveau_suggere":99.6}],"qualite_prix":{"prix_mixte_eur":9.755,"score_attendu":90.3,"avantage":1,"avantage_bas":-0.3,"avantage_haut":2.2,"meilleur_compromis":false,"rang":51}}}