{"harness": {"slug": "aide", "name": "AIDE", "vendor": "Wecoai", "description": "AIDE is tracked in this catalog as a software-engineering agent harness. Public benchmark rows use this slug on leaderboards such as SWE-bench Verified or Artificial Analysis coding-agent evals.\n\nCheck each score row's `source` and `source_url` for the authoritative snapshot. Harness-level metadata here does not invent scores \u2014 only documents what the harness is and where published results come from.", "card_summary": "Wecoai AIDE coding agent harness.", "homepage_url": "https://github.com/wecoai/aideml", "repo_url": "https://github.com/wecoai/aideml", "docs_url": null, "scaffold_code": null, "harbor_agent_name": null, "api_protocol": null, "status": "validated", "supports_rl": true, "score_count": 4, "best_success_rate": 17.12, "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": 17.12, "resolved_count": null, "total_count": null, "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": "MLE-bench All split any_medal_percentage", "source": "mle_bench_site", "source_url": "https://github.com/openai/mle-bench", "observed_at": "2024-10-08T00:00:00+00:00", "harness": {"slug": "aide", "name": "AIDE", "vendor": "Wecoai", "repo_url": "https://github.com/wecoai/aideml", "openrouter_icon_url": null}, "model": {"slug": "o1-preview", "name": "o1 preview", "vendor": "OpenAI"}, "benchmark": {"slug": "mle-bench", "name": "MLE-bench"}}, {"success_rate": 8.63, "resolved_count": null, "total_count": null, "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": "MLE-bench All split any_medal_percentage", "source": "mle_bench_site", "source_url": "https://github.com/openai/mle-bench", "observed_at": "2024-10-08T00:00:00+00:00", "harness": {"slug": "aide", "name": "AIDE", "vendor": "Wecoai", "repo_url": "https://github.com/wecoai/aideml", "openrouter_icon_url": null}, "model": {"slug": "gpt-4o-20240806", "name": "GPT-4o (2024-08-06)", "vendor": "OpenAI"}, "benchmark": {"slug": "mle-bench", "name": "MLE-bench"}}, {"success_rate": 7.56, "resolved_count": null, "total_count": null, "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": "MLE-bench All split any_medal_percentage", "source": "mle_bench_site", "source_url": "https://github.com/openai/mle-bench", "observed_at": "2024-10-08T00:00:00+00:00", "harness": {"slug": "aide", "name": "AIDE", "vendor": "Wecoai", "repo_url": "https://github.com/wecoai/aideml", "openrouter_icon_url": null}, "model": {"slug": "claude-3-5-sonnet-20240620", "name": "Claude 3.5 Sonnet (2024-06-20)", "vendor": "Anthropic"}, "benchmark": {"slug": "mle-bench", "name": "MLE-bench"}}, {"success_rate": 3.33, "resolved_count": null, "total_count": null, "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": "MLE-bench All split any_medal_percentage", "source": "mle_bench_site", "source_url": "https://github.com/openai/mle-bench", "observed_at": "2024-10-08T00:00:00+00:00", "harness": {"slug": "aide", "name": "AIDE", "vendor": "Wecoai", "repo_url": "https://github.com/wecoai/aideml", "openrouter_icon_url": null}, "model": {"slug": "llama-3-1-405b-instruct", "name": "Llama 3.1 405b Instruct", "vendor": "Meta"}, "benchmark": {"slug": "mle-bench", "name": "MLE-bench"}}], "benchmarks": [{"slug": "mle-bench", "name": "MLE-bench", "category": "eval"}]}