热度58.9▼ 1.2
SWE-bench Verified
swe-bench / resolved · 180 行结果
| 排名 | 模型 | 子项 | 分数 | Metric | 时间 |
|---|---|---|---|---|---|
| #1 | Claude 4.5 Opus Anthropiclive-SWE-agent + Claude 4.5 Opus medium (20251101) | SWE-bench Verified swe-bench | 79.2% | resolved | 2025-12-15 |
| #2 | Claude 4.5 Opus AnthropicSonar Foundation Agent + Claude 4.5 Opus | SWE-bench Verified swe-bench | 79.2% | resolved | 2025-12-05 |
| #3 | Doubao-Seed-Code ByteDanceTRAE + Doubao-Seed-Code | SWE-bench Verified swe-bench | 78.8% | resolved | 2025-09-28 |
| #4 | Gemini 3 Pro Preview Google DeepMindlive-SWE-agent + Gemini 3 Pro Preview (2025-11-18) | SWE-bench Verified swe-bench | 77.4% | resolved | 2025-11-20 |
| #5 | Multiple AnthropicAtlassian Rovo Dev (2025-09-02) | SWE-bench Verified swe-bench | 76.8% | resolved | 2025-09-02 |
| #6 | Claude 4 Sonnet AnthropicEPAM AI/Run Developer Agent v20250719 + Claude 4 Sonnet | SWE-bench Verified swe-bench | 76.8% | resolved | 2025-08-04 |
| #7 | Claude 4.5 Opus Anthropicmini-SWE-agent + Claude 4.5 Opus (high) | SWE-bench Verified swe-bench | 76.8% | resolved | 2026-02-17 |
| #8 | Multiple AnthropicACoder | SWE-bench Verified swe-bench | 76.4% | resolved | 2025-08-19 |
| #9 | Gemini 3 Flash Google DeepMindmini-SWE-agent + Gemini 3 Flash (high) | SWE-bench Verified swe-bench | 75.8% | resolved | 2026-02-17 |
| #10 | MiniMax M2.5 Minimaxmini-SWE-agent + MiniMax M2.5 (high) | SWE-bench Verified swe-bench | 75.8% | resolved | 2026-02-17 |
| #11 | Multiple OpenAIWarp | SWE-bench Verified swe-bench | 75.6% | resolved | 2025-09-01 |
| #12 | Claude 4.6 Opus Anthropicmini-SWE-agent + Claude 4.6 Opus | SWE-bench Verified swe-bench | 75.6% | resolved | 2026-02-17 |
| #13 | Multiple AnthropicTRAE + Claude Sonnet 4 + Opus 4 + Sonnet 3.7 + Gemini 2.5 Pro | SWE-bench Verified swe-bench | 75.2% | resolved | 2025-06-12 |
| #14 | Claude Sonnet 4 AnthropicHarness AI | SWE-bench Verified swe-bench | 74.8% | resolved | 2025-07-31 |
| #15 | Claude 4.5 Sonnet AnthropicSonar Foundation Agent + Claude 4.5 Sonnet | SWE-bench Verified swe-bench | 74.8% | resolved | 2025-11-03 |
| #16 | Claude 4 Sonnet AnthropicLingxi-v1.5_claude-4-sonnet-20250514 | SWE-bench Verified swe-bench | 74.6% | resolved | 2025-07-20 |
| #17 | Multiple AnthropicJoyCode + Claude 4 Sonnet + GPT-4.1 | SWE-bench Verified swe-bench | 74.6% | resolved | 2025-09-15 |
| #18 | Multiple AnthropicRefact.ai Agent + Claude 4 Sonnet + o4-mini | SWE-bench Verified swe-bench | 74.4% | resolved | 2025-06-03 |
| #19 | GPT-5 OpenAIPrometheus-v1.2.1 + GPT-5 | SWE-bench Verified swe-bench | 74.4% | resolved | 2025-10-15 |
| #20 | Claude 4.5 Opus Anthropicmini-SWE-agent + Claude 4.5 Opus (20251101) (medium) | SWE-bench Verified swe-bench | 74.4% | resolved | 2025-11-24 |
| #21 | Gemini 3 Pro Preview Google DeepMindmini-SWE-agent + Gemini 3 Pro Preview (2025-11-18) | SWE-bench Verified swe-bench | 74.2% | resolved | 2025-11-18 |
| #22 | Multiple AnthropicSalesforce AI Research SAGE (OpenHands) | SWE-bench Verified swe-bench | 73.8% | resolved | 2025-11-03 |
| #23 | Claude 4 Opus AnthropicTools + Claude 4 Opus (2025-05-22) | SWE-bench Verified swe-bench | 73.2% | resolved | 2025-05-22 |
| #24 | Multiple AnthropicSalesforce AI Research SAGE (bash-only) | SWE-bench Verified swe-bench | 73% | resolved | 2025-10-21 |
| #25 | GPT 5.2 Codex OpenAImini-SWE-agent + GPT 5.2 Codex | SWE-bench Verified swe-bench | 72.8% | resolved | 2026-02-19 |
180 条结果 · 当前 1-25
1 / 8