{
 "id": "capy",
 "nom": "Capy",
 "url": "https://quelleia.com/outils/capy/",
 "categorie": "agent_classement",
 "type": "autre",
 "type_nom": "autre agent",
 "soumis_par": "Capy",
 "lien_declare": "https://capy.ai",
 "lien_source": "https://huggingface.co/datasets/harborframework/terminal-bench-2-leaderboard/blob/main/submissions/terminal-bench/2.0/Capy__GPT-5.5/metadata.yaml",
 "depot": null,
 "note": null,
 "ecritures": [
  "Capy"
 ],
 "resultats_mesures": [
  {
   "benchmark": "terminal_bench",
   "modele": "gpt-5-5",
   "valeur": 83.15,
   "unite": "%",
   "sortie_modele": "2026-04-23",
   "source": "https://www.tbench.ai/leaderboard/terminal-bench/2.0",
   "rang": 2,
   "sur": 4,
   "meilleur_outil_grand_public": {
    "nom": "Codex CLI",
    "valeur": 82.25
   },
   "ecart_avec_cet_outil": 0.9,
   "comparateur": "https://quelleia.com/outils/harnais/?benchmark=terminal_bench&modele=gpt-5-5#terminal_bench--gpt-5-5"
  },
  {
   "benchmark": "terminal_bench",
   "modele": "claude-opus-4-6",
   "valeur": 75.28,
   "unite": "%",
   "sortie_modele": "2026-02-05",
   "source": "https://www.tbench.ai/leaderboard/terminal-bench/2.0",
   "rang": 2,
   "sur": 10,
   "meilleur_outil_grand_public": {
    "nom": "Droid (Factory)",
    "valeur": 69.89
   },
   "ecart_avec_cet_outil": 5.39,
   "comparateur": "https://quelleia.com/outils/harnais/?benchmark=terminal_bench&modele=claude-opus-4-6#terminal_bench--claude-opus-4-6"
  }
 ],
 "retirees": [],
 "ce_qu_on_ne_sait_pas": [
  "Comment l'installer, ce qu'il coûte, sur quels systèmes il tourne : sa soumission au classement ne le dit pas, et nous n'avons rien relevé d'autre.",
  "Sa soumission ne cite aucun dépôt public."
 ],
 "sources": [
  "https://huggingface.co/datasets/harborframework/terminal-bench-2-leaderboard/blob/main/submissions/terminal-bench/2.0/Capy__GPT-5.5/metadata.yaml",
  "https://www.tbench.ai/leaderboard/terminal-bench/2.0"
 ]
}