{"model": {"slug": "gemini-3-0-pro", "name": "Gemini 3.0 Pro", "vendor": "Google", "model_family": "gemini-3-0-pro", "parameter_count_b": null, "context_length": 1000000, "homepage_url": null, "hf_model_id": null, "openrouter_id": null, "card_summary": "Google Gemini 3.0 Pro on SWE-bench Verified leaderboards.", "is_rl_checkpoint": false, "score_count": 1, "best_success_rate": 85.5, "input_price_per_million": null, "output_price_per_million": null}, "scores": [{"success_rate": 85.5, "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": "HAL reliability composite (mean of 4 dimension aggregates)", "source": "hal_reliability_site", "source_url": "https://hal.cs.princeton.edu/reliability/benchmark/taubench_airline_original/", "observed_at": null, "harness": {"slug": "hal-generalist-agent", "name": "HAL Generalist Agent", "vendor": "Princeton PLI", "repo_url": null, "openrouter_icon_url": null}, "model": {"slug": "gemini-3-0-pro", "name": "Gemini 3.0 Pro", "vendor": "Google"}, "benchmark": {"slug": "hal-reliability-tau-bench-airline-original", "name": "HAL Reliability \u2014 \u03c4-bench Airline (original)"}}], "harnesses": [{"slug": "hal-generalist-agent", "name": "HAL Generalist Agent", "vendor": "Princeton PLI", "repo_url": "https://github.com/princeton-pli/hal-harness", "openrouter_icon_url": null}]}