{"harness": {"slug": "prometheus-v1-2", "name": "Prometheus V1 2", "vendor": "EuniAI", "description": "Earlier Prometheus v1.2 submission on SWE-bench Verified. Compare against v1.2.1 only when models and task protocol align \u2014 version bumps may change prompts or tool interfaces.\n\n`GAIR-NLP/Prometheus` on GitHub. `swe_bench_site` ingest.", "card_summary": "Prometheus v1.2 agent \u2014 earlier Verified submission build.", "homepage_url": "https://github.com/GAIR-NLP/Prometheus", "repo_url": "https://github.com/GAIR-NLP/Prometheus", "docs_url": null, "scaffold_code": null, "harbor_agent_name": null, "api_protocol": null, "status": "validated", "supports_rl": true, "score_count": 1, "best_success_rate": 71.2, "github_stars": null, "popularity_tier": null, "hrl_score": null, "hrl_rank": null, "catalog_token_volume": null, "openrouter_icon_url": null, "leaderboard_icon_url": "https://www.swebench.com/img/logos/20250929_Prometheus_v1.2_gpt5.jpg", "aa_coding_index": null, "aa_mean_cost_usd": null, "aa_mean_total_tokens": null}, "scores": [{"success_rate": 71.2, "resolved_count": 356, "total_count": 500, "cost_usd": null, "input_tokens": null, "output_tokens": null, "latency_ms": null, "temperature": null, "max_turns": null, "context_budget_tokens": null, "harness_version": null, "trial_count": null, "is_harnessrl_measured": false, "eval_protocol": "SWE-bench Verified public submission", "source": "swe_bench_site", "source_url": "https://euni.ai/", "observed_at": "2025-09-29T00:00:00+00:00", "harness": {"slug": "prometheus-v1-2", "name": "Prometheus V1 2", "vendor": "EuniAI", "repo_url": "https://github.com/GAIR-NLP/Prometheus", "openrouter_icon_url": null}, "model": {"slug": "gpt-5", "name": "GPT-5", "vendor": "openai"}, "benchmark": {"slug": "swe-bench-verified", "name": "SWE-bench Verified"}}], "benchmarks": [{"slug": "swe-bench-verified", "name": "SWE-bench Verified", "category": "eval"}]}