{"harness": {"slug": "masai", "name": "Masai", "vendor": "Masai", "description": "Masai appears as a harness label on SWE-bench Verified community submissions. Typically a product or research coding agent evaluated on the verified 500-issue split.\n\n`swe_bench_site` ingest. Read model and source URL per row.", "card_summary": "Masai agent \u2014 SWE-bench Verified community submission.", "homepage_url": "https://www.swebench.com/verified", "repo_url": null, "docs_url": null, "scaffold_code": null, "harbor_agent_name": null, "api_protocol": null, "status": "validated", "supports_rl": true, "score_count": 2, "best_success_rate": 32.6, "github_stars": null, "popularity_tier": null, "hrl_score": null, "hrl_rank": null, "catalog_token_volume": null, "openrouter_icon_url": null, "leaderboard_icon_url": null, "aa_coding_index": null, "aa_mean_cost_usd": null, "aa_mean_total_tokens": null}, "scores": [{"success_rate": 32.6, "resolved_count": 163, "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://masai-dev-agent.github.io/", "observed_at": "2024-06-12T00:00:00+00:00", "harness": {"slug": "masai", "name": "Masai", "vendor": "Masai", "repo_url": null, "openrouter_icon_url": null}, "model": {"slug": "gpt-4o-20240612", "name": "GPT-4o (2024-06-12)", "vendor": "OpenAI"}, "benchmark": {"slug": "swe-bench-verified", "name": "SWE-bench Verified"}}, {"success_rate": 27.33, "resolved_count": 82, "total_count": 300, "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 Lite public submission", "source": "swe_bench_site", "source_url": "https://masai-dev-agent.github.io/", "observed_at": "2024-06-12T00:00:00+00:00", "harness": {"slug": "masai", "name": "Masai", "vendor": "Masai", "repo_url": null, "openrouter_icon_url": null}, "model": {"slug": "gpt-4o-20240513", "name": "GPT-4o (2024-05-13)", "vendor": "OpenAI"}, "benchmark": {"slug": "swe-bench-lite", "name": "SWE-bench Lite"}}], "benchmarks": [{"slug": "swe-bench-lite", "name": "SWE-bench Lite", "category": "eval"}, {"slug": "swe-bench-verified", "name": "SWE-bench Verified", "category": "eval"}]}