{"model": {"slug": "deepseek-v3-2-reasoner", "name": "DeepSeek V3.2 Reasoner", "vendor": "DeepSeek", "model_family": "deepseek-v3-2-reasoner", "parameter_count_b": null, "context_length": 128000, "homepage_url": null, "hf_model_id": null, "openrouter_id": null, "card_summary": "DeepSeek DeepSeek V3.2 Reasoner on SWE-bench Verified leaderboards.", "is_rl_checkpoint": false, "score_count": 7, "best_success_rate": 71.6, "input_price_per_million": null, "output_price_per_million": null}, "scores": [{"success_rate": 71.6, "resolved_count": null, "total_count": null, "cost_usd": 0.16, "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": "OpenHands SDK v1.8.3", "source": "openhands_index", "source_url": "https://index.openhands.dev/", "observed_at": "2026-07-31T11:38:58.577710", "harness": {"slug": "openhands-sdk", "name": "OpenHands SDK", "vendor": "All Hands AI", "repo_url": null, "openrouter_icon_url": null}, "model": {"slug": "deepseek-v3-2-reasoner", "name": "DeepSeek V3.2 Reasoner", "vendor": "DeepSeek"}, "benchmark": {"slug": "openhands-swe-bench", "name": "SWE-bench (OpenHands Index)"}}, {"success_rate": 60.0, "resolved_count": 300, "total_count": 500, "cost_usd": 0.0280763028, "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 bash-only (mini-SWE-agent)", "source": "swe_bench_site", "source_url": "https://www.swebench.com/verified", "observed_at": "2025-12-01T00:00:00+00:00", "harness": {"slug": "mini-swe-agent", "name": "mini-SWE-agent", "vendor": "anthropic", "repo_url": null, "openrouter_icon_url": null}, "model": {"slug": "deepseek-v3-2-reasoner", "name": "DeepSeek V3.2 Reasoner", "vendor": "DeepSeek"}, "benchmark": {"slug": "swe-bench-bash-only", "name": "SWE-bench Verified (bash-only)"}}, {"success_rate": 60.0, "resolved_count": 300, "total_count": 500, "cost_usd": 0.0280763028, "input_tokens": null, "output_tokens": null, "latency_ms": null, "temperature": null, "max_turns": null, "context_budget_tokens": null, "harness_version": null, "trial_count": 46, "is_harnessrl_measured": false, "eval_protocol": "SWE-bench Verified public submission", "source": "swe_bench_site", "source_url": "https://mini-swe-agent.com/latest/", "observed_at": "2025-12-01T00:00:00+00:00", "harness": {"slug": "mini-swe-agent", "name": "mini-SWE-agent", "vendor": "anthropic", "repo_url": null, "openrouter_icon_url": null}, "model": {"slug": "deepseek-v3-2-reasoner", "name": "DeepSeek V3.2 Reasoner", "vendor": "DeepSeek"}, "benchmark": {"slug": "swe-bench-verified", "name": "SWE-bench Verified"}}, {"success_rate": 53.6, "resolved_count": null, "total_count": null, "cost_usd": 0.12, "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": "OpenHands SDK v1.8.3", "source": "openhands_index", "source_url": "https://index.openhands.dev/", "observed_at": "2026-07-31T11:38:58.577710", "harness": {"slug": "openhands-sdk", "name": "OpenHands SDK", "vendor": "All Hands AI", "repo_url": null, "openrouter_icon_url": null}, "model": {"slug": "deepseek-v3-2-reasoner", "name": "DeepSeek V3.2 Reasoner", "vendor": "DeepSeek"}, "benchmark": {"slug": "swt-bench", "name": "SWT-Bench"}}, {"success_rate": 50.3, "resolved_count": null, "total_count": null, "cost_usd": 0.06, "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": "OpenHands SDK v1.8.3", "source": "openhands_index", "source_url": "https://index.openhands.dev/", "observed_at": "2026-07-31T11:38:58.577710", "harness": {"slug": "openhands-sdk", "name": "OpenHands SDK", "vendor": "All Hands AI", "repo_url": null, "openrouter_icon_url": null}, "model": {"slug": "deepseek-v3-2-reasoner", "name": "DeepSeek V3.2 Reasoner", "vendor": "DeepSeek"}, "benchmark": {"slug": "gaia", "name": "GAIA"}}, {"success_rate": 27.9, "resolved_count": null, "total_count": null, "cost_usd": 0.19, "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": "OpenHands SDK v1.8.3", "source": "openhands_index", "source_url": "https://index.openhands.dev/", "observed_at": "2026-07-31T11:38:58.577710", "harness": {"slug": "openhands-sdk", "name": "OpenHands SDK", "vendor": "All Hands AI", "repo_url": null, "openrouter_icon_url": null}, "model": {"slug": "deepseek-v3-2-reasoner", "name": "DeepSeek V3.2 Reasoner", "vendor": "DeepSeek"}, "benchmark": {"slug": "swe-bench-multimodal", "name": "SWE-bench Multimodal (OpenHands Index)"}}, {"success_rate": 25.0, "resolved_count": null, "total_count": null, "cost_usd": 0.57, "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": "OpenHands SDK v1.8.3", "source": "openhands_index", "source_url": "https://laminar.sh/shared/evals/df76145a-e319-468e-ae2e-07f71208496f", "observed_at": "2026-07-31T11:38:58.577710", "harness": {"slug": "openhands-sdk", "name": "OpenHands SDK", "vendor": "All Hands AI", "repo_url": null, "openrouter_icon_url": null}, "model": {"slug": "deepseek-v3-2-reasoner", "name": "DeepSeek V3.2 Reasoner", "vendor": "DeepSeek"}, "benchmark": {"slug": "commit0", "name": "Commit0"}}], "harnesses": [{"slug": "openhands-sdk", "name": "OpenHands SDK", "vendor": "All Hands AI", "repo_url": "https://github.com/All-Hands-AI/OpenHands", "openrouter_icon_url": "https://www.all-hands.dev/favicon.ico"}, {"slug": "mini-swe-agent", "name": "mini-SWE-agent", "vendor": "anthropic", "repo_url": "https://github.com/SWE-agent/mini-swe-agent", "openrouter_icon_url": null}]}