{"harness": {"slug": "cursor-cli", "name": "Cursor CLI", "vendor": "Cursor", "description": "Cursor CLI exposes Cursor's coding agent as a terminal command: repository indexing, multi-file edits, and tool use aimed at professional IDE-adjacent workflows. Leaderboard entries typically pair Cursor's harness with frontier models on SWE-bench Verified and Artificial Analysis benchmarks.\n\nThis is a product harness (not a minimal research scaffold). Scores reflect Cursor's prompts, tools, and retrieval stack in addition to the backing model. Compare like-with-like: Cursor CLI vs Claude Code vs Codex, not vs bash-only mini-SWE-agent baselines.\n\nScores in this catalog are ingested from `artificialanalysis` and `swe_bench_site`; each row links to its source snapshot.", "card_summary": "Cursor's CLI agent for repo-aware edits, codebase search, and shell workflows.", "homepage_url": "https://cursor.com", "repo_url": "https://github.com/getcursor/cursor", "docs_url": null, "scaffold_code": null, "harbor_agent_name": null, "api_protocol": null, "status": "validated", "supports_rl": true, "score_count": 14, "best_success_rate": 79.33, "github_stars": null, "popularity_tier": null, "hrl_score": 48.0, "hrl_rank": 11, "catalog_token_volume": 0, "openrouter_icon_url": "https://cursor.com/apple-touch-icon.png", "leaderboard_icon_url": null, "aa_coding_index": 0.47086936005984537, "aa_mean_cost_usd": 0.08505141359918204, "aa_mean_total_tokens": 3705076}, "scores": [{"success_rate": 79.33, "resolved_count": null, "total_count": null, "cost_usd": 0.3013258426966292, "input_tokens": 12570445, "output_tokens": 4734062, "latency_ms": null, "temperature": null, "max_turns": null, "context_budget_tokens": null, "harness_version": null, "trial_count": 445, "is_harnessrl_measured": false, "eval_protocol": null, "source": "tbench_official", "source_url": "https://github.com/harbor-framework/terminal-bench-2-1/pull/86", "observed_at": "2026-07-09T00:00:00+00:00", "harness": {"slug": "cursor-cli", "name": "Cursor CLI", "vendor": "Cursor", "repo_url": "https://github.com/getcursor/cursor", "openrouter_icon_url": "https://cursor.com/apple-touch-icon.png"}, "model": {"slug": "grok-4-5", "name": "GROK 4.5", "vendor": "x-ai"}, "benchmark": {"slug": "terminal-bench-2-1-official", "name": "Terminal-Bench 2.1 (official)"}}, {"success_rate": 76.4, "resolved_count": null, "total_count": null, "cost_usd": 1.9980312342194944, "input_tokens": 2043912, "output_tokens": 11032, "latency_ms": null, "temperature": null, "max_turns": null, "context_budget_tokens": null, "harness_version": null, "trial_count": null, "is_harnessrl_measured": false, "eval_protocol": "Artificial Analysis Coding Agent Index component (pass@1)", "source": "artificialanalysis", "source_url": "https://artificialanalysis.ai/agents/coding-agents", "observed_at": "2026-08-27T21:54:11.189561+00:00", "harness": {"slug": "cursor-cli", "name": "Cursor CLI", "vendor": "Cursor", "repo_url": "https://github.com/getcursor/cursor", "openrouter_icon_url": "https://cursor.com/apple-touch-icon.png"}, "model": {"slug": "gpt-5-5-medium", "name": "GPT-5.5 (medium)", "vendor": "OpenAI"}, "benchmark": {"slug": "terminal-bench-v2-1", "name": "Terminal-Bench v2.1 (AA component)"}}, {"success_rate": 74.53, "resolved_count": null, "total_count": null, "cost_usd": 2.72490571101738, "input_tokens": 2855520, "output_tokens": 17872, "latency_ms": null, "temperature": null, "max_turns": null, "context_budget_tokens": null, "harness_version": null, "trial_count": null, "is_harnessrl_measured": false, "eval_protocol": "Artificial Analysis Coding Agent Index component (pass@1)", "source": "artificialanalysis", "source_url": "https://artificialanalysis.ai/agents/coding-agents", "observed_at": "2026-08-27T21:54:11.189561+00:00", "harness": {"slug": "cursor-cli", "name": "Cursor CLI", "vendor": "Cursor", "repo_url": "https://github.com/getcursor/cursor", "openrouter_icon_url": "https://cursor.com/apple-touch-icon.png"}, "model": {"slug": "claude-4-7-opus-medium", "name": "Claude 4.7 Opus (medium)", "vendor": "Anthropic"}, "benchmark": {"slug": "terminal-bench-v2-1", "name": "Terminal-Bench v2.1 (AA component)"}}, {"success_rate": 68.24, "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": "SWE-rebench resolved rate (best rolling window)", "source": "swe_rebench_site", "source_url": "https://swe-rebench.com/", "observed_at": null, "harness": {"slug": "cursor-cli", "name": "Cursor CLI", "vendor": "Cursor", "repo_url": "https://github.com/getcursor/cursor", "openrouter_icon_url": "https://cursor.com/apple-touch-icon.png"}, "model": {"slug": "swe-rebench-agent-product", "name": "SWE-rebench agent (product)", "vendor": null}, "benchmark": {"slug": "swe-rebench", "name": "SWE-rebench"}}, {"success_rate": 67.79, "resolved_count": null, "total_count": null, "cost_usd": 0.5573437730061347, "input_tokens": 2153033, "output_tokens": 20146, "latency_ms": null, "temperature": null, "max_turns": null, "context_budget_tokens": null, "harness_version": null, "trial_count": null, "is_harnessrl_measured": false, "eval_protocol": "Artificial Analysis Coding Agent Index component (pass@1)", "source": "artificialanalysis", "source_url": "https://artificialanalysis.ai/agents/coding-agents", "observed_at": "2026-08-27T21:54:11.189561+00:00", "harness": {"slug": "cursor-cli", "name": "Cursor CLI", "vendor": "Cursor", "repo_url": "https://github.com/getcursor/cursor", "openrouter_icon_url": "https://cursor.com/apple-touch-icon.png"}, "model": {"slug": "composer-2-5-fast", "name": "Composer 2 5 Fast", "vendor": "Cursor"}, "benchmark": {"slug": "terminal-bench-v2-1", "name": "Terminal-Bench v2.1 (AA component)"}}, {"success_rate": 67.79, "resolved_count": null, "total_count": null, "cost_usd": 0.08505141359918204, "input_tokens": 1872461, "output_tokens": 18453, "latency_ms": null, "temperature": null, "max_turns": null, "context_budget_tokens": null, "harness_version": null, "trial_count": null, "is_harnessrl_measured": false, "eval_protocol": "Artificial Analysis Coding Agent Index component (pass@1)", "source": "artificialanalysis", "source_url": "https://artificialanalysis.ai/agents/coding-agents", "observed_at": "2026-08-27T21:54:11.189561+00:00", "harness": {"slug": "cursor-cli", "name": "Cursor CLI", "vendor": "Cursor", "repo_url": "https://github.com/getcursor/cursor", "openrouter_icon_url": "https://cursor.com/apple-touch-icon.png"}, "model": {"slug": "composer-2-5", "name": "Composer 2 5", "vendor": "Cursor"}, "benchmark": {"slug": "terminal-bench-v2-1", "name": "Terminal-Bench v2.1 (AA component)"}}, {"success_rate": 37.17, "resolved_count": null, "total_count": null, "cost_usd": 1.9980312342194944, "input_tokens": 2043912, "output_tokens": 11032, "latency_ms": null, "temperature": null, "max_turns": null, "context_budget_tokens": null, "harness_version": null, "trial_count": null, "is_harnessrl_measured": false, "eval_protocol": "Artificial Analysis Coding Agent Index component (pass@1)", "source": "artificialanalysis", "source_url": "https://artificialanalysis.ai/agents/coding-agents", "observed_at": "2026-08-27T21:54:11.189561+00:00", "harness": {"slug": "cursor-cli", "name": "Cursor CLI", "vendor": "Cursor", "repo_url": "https://github.com/getcursor/cursor", "openrouter_icon_url": "https://cursor.com/apple-touch-icon.png"}, "model": {"slug": "gpt-5-5-medium", "name": "GPT-5.5 (medium)", "vendor": "OpenAI"}, "benchmark": {"slug": "deepswe", "name": "DeepSWE (AA component)"}}, {"success_rate": 33.87, "resolved_count": null, "total_count": null, "cost_usd": 2.72490571101738, "input_tokens": 2855520, "output_tokens": 17872, "latency_ms": null, "temperature": null, "max_turns": null, "context_budget_tokens": null, "harness_version": null, "trial_count": null, "is_harnessrl_measured": false, "eval_protocol": "Artificial Analysis Coding Agent Index component (pass@1)", "source": "artificialanalysis", "source_url": "https://artificialanalysis.ai/agents/coding-agents", "observed_at": "2026-08-27T21:54:11.189561+00:00", "harness": {"slug": "cursor-cli", "name": "Cursor CLI", "vendor": "Cursor", "repo_url": "https://github.com/getcursor/cursor", "openrouter_icon_url": "https://cursor.com/apple-touch-icon.png"}, "model": {"slug": "claude-4-7-opus-medium", "name": "Claude 4.7 Opus (medium)", "vendor": "Anthropic"}, "benchmark": {"slug": "swe-atlas-qna", "name": "SWE-Atlas-QnA (AA component)"}}, {"success_rate": 31.56, "resolved_count": null, "total_count": null, "cost_usd": 2.72490571101738, "input_tokens": 2855520, "output_tokens": 17872, "latency_ms": null, "temperature": null, "max_turns": null, "context_budget_tokens": null, "harness_version": null, "trial_count": null, "is_harnessrl_measured": false, "eval_protocol": "Artificial Analysis Coding Agent Index component (pass@1)", "source": "artificialanalysis", "source_url": "https://artificialanalysis.ai/agents/coding-agents", "observed_at": "2026-08-27T21:54:11.189561+00:00", "harness": {"slug": "cursor-cli", "name": "Cursor CLI", "vendor": "Cursor", "repo_url": "https://github.com/getcursor/cursor", "openrouter_icon_url": "https://cursor.com/apple-touch-icon.png"}, "model": {"slug": "claude-4-7-opus-medium", "name": "Claude 4.7 Opus (medium)", "vendor": "Anthropic"}, "benchmark": {"slug": "deepswe", "name": "DeepSWE (AA component)"}}, {"success_rate": 31.18, "resolved_count": null, "total_count": null, "cost_usd": 0.5573437730061347, "input_tokens": 2153033, "output_tokens": 20146, "latency_ms": null, "temperature": null, "max_turns": null, "context_budget_tokens": null, "harness_version": null, "trial_count": null, "is_harnessrl_measured": false, "eval_protocol": "Artificial Analysis Coding Agent Index component (pass@1)", "source": "artificialanalysis", "source_url": "https://artificialanalysis.ai/agents/coding-agents", "observed_at": "2026-08-27T21:54:11.189561+00:00", "harness": {"slug": "cursor-cli", "name": "Cursor CLI", "vendor": "Cursor", "repo_url": "https://github.com/getcursor/cursor", "openrouter_icon_url": "https://cursor.com/apple-touch-icon.png"}, "model": {"slug": "composer-2-5-fast", "name": "Composer 2 5 Fast", "vendor": "Cursor"}, "benchmark": {"slug": "swe-atlas-qna", "name": "SWE-Atlas-QnA (AA component)"}}, {"success_rate": 31.18, "resolved_count": null, "total_count": null, "cost_usd": 0.08505141359918204, "input_tokens": 1872461, "output_tokens": 18453, "latency_ms": null, "temperature": null, "max_turns": null, "context_budget_tokens": null, "harness_version": null, "trial_count": null, "is_harnessrl_measured": false, "eval_protocol": "Artificial Analysis Coding Agent Index component (pass@1)", "source": "artificialanalysis", "source_url": "https://artificialanalysis.ai/agents/coding-agents", "observed_at": "2026-08-27T21:54:11.189561+00:00", "harness": {"slug": "cursor-cli", "name": "Cursor CLI", "vendor": "Cursor", "repo_url": "https://github.com/getcursor/cursor", "openrouter_icon_url": "https://cursor.com/apple-touch-icon.png"}, "model": {"slug": "composer-2-5", "name": "Composer 2 5", "vendor": "Cursor"}, "benchmark": {"slug": "swe-atlas-qna", "name": "SWE-Atlas-QnA (AA component)"}}, {"success_rate": 27.69, "resolved_count": null, "total_count": null, "cost_usd": 1.9980312342194944, "input_tokens": 2043912, "output_tokens": 11032, "latency_ms": null, "temperature": null, "max_turns": null, "context_budget_tokens": null, "harness_version": null, "trial_count": null, "is_harnessrl_measured": false, "eval_protocol": "Artificial Analysis Coding Agent Index component (pass@1)", "source": "artificialanalysis", "source_url": "https://artificialanalysis.ai/agents/coding-agents", "observed_at": "2026-08-27T21:54:11.189561+00:00", "harness": {"slug": "cursor-cli", "name": "Cursor CLI", "vendor": "Cursor", "repo_url": "https://github.com/getcursor/cursor", "openrouter_icon_url": "https://cursor.com/apple-touch-icon.png"}, "model": {"slug": "gpt-5-5-medium", "name": "GPT-5.5 (medium)", "vendor": "OpenAI"}, "benchmark": {"slug": "swe-atlas-qna", "name": "SWE-Atlas-QnA (AA component)"}}, {"success_rate": 15.93, "resolved_count": null, "total_count": null, "cost_usd": 0.5573437730061347, "input_tokens": 2153033, "output_tokens": 20146, "latency_ms": null, "temperature": null, "max_turns": null, "context_budget_tokens": null, "harness_version": null, "trial_count": null, "is_harnessrl_measured": false, "eval_protocol": "Artificial Analysis Coding Agent Index component (pass@1)", "source": "artificialanalysis", "source_url": "https://artificialanalysis.ai/agents/coding-agents", "observed_at": "2026-08-27T21:54:11.189561+00:00", "harness": {"slug": "cursor-cli", "name": "Cursor CLI", "vendor": "Cursor", "repo_url": "https://github.com/getcursor/cursor", "openrouter_icon_url": "https://cursor.com/apple-touch-icon.png"}, "model": {"slug": "composer-2-5-fast", "name": "Composer 2 5 Fast", "vendor": "Cursor"}, "benchmark": {"slug": "deepswe", "name": "DeepSWE (AA component)"}}, {"success_rate": 15.93, "resolved_count": null, "total_count": null, "cost_usd": 0.08505141359918204, "input_tokens": 1872461, "output_tokens": 18453, "latency_ms": null, "temperature": null, "max_turns": null, "context_budget_tokens": null, "harness_version": null, "trial_count": null, "is_harnessrl_measured": false, "eval_protocol": "Artificial Analysis Coding Agent Index component (pass@1)", "source": "artificialanalysis", "source_url": "https://artificialanalysis.ai/agents/coding-agents", "observed_at": "2026-08-27T21:54:11.189561+00:00", "harness": {"slug": "cursor-cli", "name": "Cursor CLI", "vendor": "Cursor", "repo_url": "https://github.com/getcursor/cursor", "openrouter_icon_url": "https://cursor.com/apple-touch-icon.png"}, "model": {"slug": "composer-2-5", "name": "Composer 2 5", "vendor": "Cursor"}, "benchmark": {"slug": "deepswe", "name": "DeepSWE (AA component)"}}], "benchmarks": [{"slug": "deepswe", "name": "DeepSWE (AA component)", "category": "eval"}, {"slug": "swe-atlas-qna", "name": "SWE-Atlas-QnA (AA component)", "category": "eval"}, {"slug": "swe-rebench", "name": "SWE-rebench", "category": "eval"}, {"slug": "terminal-bench-2-1-official", "name": "Terminal-Bench 2.1 (official)", "category": "eval"}, {"slug": "terminal-bench-v2-1", "name": "Terminal-Bench v2.1 (AA component)", "category": "eval"}]}