SWE-Bench Verified
A verified subset of 500 software engineering problems from real GitHub issues, validated by human annotators for evaluating language models' ability to resolve real-world coding issues by generating patches for Python codebases.
Claude Mythos Preview from Anthropic currently leads the SWE-Bench Verified leaderboard with a score of 0.939 across 97 evaluated AI models.
What SWE-Bench Verified measures
SWE-Bench Verified is a text benchmark that evaluates large language models on frontend development, reasoning, and code tasks. LLM Stats tracks 97 models on this benchmark, with a maximum possible score of 1. Current average across reported models is 0.7, with the leader reaching 0.9.
Compare leaders on the best AI for frontend development, best AI for reasoning and best AI for code leaderboards.
Publication
- Paper
- SWE-bench: Can Language Models Resolve Real-World GitHub Issues?
- Authors
- Carlos E. Jimenez, John Yang, Alexander Wettig, Shunyu Yao, and 3 others
- Published
- arXiv
- 2310.06770
Abstract
Language models have outpaced our ability to evaluate them effectively, but for their future development it is essential to study the frontier of their capabilities. We find real-world software engineering to be a rich, sustainable, and challenging testbed for evaluating the next generation of language models. To this end, we introduce SWE-bench, an evaluation framework consisting of $2,294$ software engineering problems drawn from real GitHub issues and corresponding pull requests across $12$ popular Python repositories. Given a codebase along with a description of an issue to be resolved, a language model is tasked with editing the codebase to address the issue. Resolving issues in SWE-bench frequently requires understanding and coordinating changes across multiple functions, classes, and even files simultaneously, calling for models to interact with execution environments, process extremely long contexts and perform complex reasoning that goes far beyond traditional code generation tasks. Our evaluations show that both state-of-the-art proprietary models and our fine-tuned model SWE-Llama can resolve only the simplest issues. The best-performing model, Claude 2, is able to solve a mere $1.96$% of the issues. Advances on SWE-bench represent steps towards LMs that are more practical, intelligent, and autonomous.
Claude Mythos Preview leads with 93.9%, followed by Claude Opus 4.8 at 88.6% and Claude Opus 4.7 at 87.6%.
Progress Over Time
Interactive timeline showing model performance evolution on SWE-Bench Verified
SWE-Bench Verified Leaderboard
| Context | Cost | License | ||||
|---|---|---|---|---|---|---|
| 1 | Anthropic | — | — | — | ||
| 2 | Anthropic | — | 1.0M | $5.00 / $25.00 | ||
| 3 | Anthropic | — | 1.0M | $5.00 / $25.00 | ||
| 4 | Anthropic | — | — | — | ||
| 5 | Anthropic | — | 1.0M | $5.00 / $25.00 | ||
| 6 |  Google | — | 1.0M | $2.50 / $15.00 | ||
| 6 |  DeepSeek | 1.6T | 1.0M | $1.74 / $3.48 | ||
| 8 |  MiniMax M3New MiniMax | — | 1.0M | $0.60 / $2.40 | ||
| 9 |  Alibaba Cloud / Qwen Team | — | 1.0M | $1.25 / $3.75 | ||
| 10 | MiniMax | 230B | 1.0M | $0.30 / $1.20 | ||
| 10 |  Moonshot AI | 1.0T | 262K | $0.95 / $4.00 | ||
| 12 |  OpenAI | — | 400K | $1.75 / $14.00 | ||
| 13 | Anthropic | — | 200K | $3.00 / $15.00 | ||
| 14 | DeepSeek | 284B | 1.0M | $0.14 / $0.28 | ||
| 15 | Alibaba Cloud / Qwen Team | — | 1.0M | $0.50 / $3.00 | ||
| 16 | Google | — | 1.0M | $0.50 / $3.00 | ||
| 16 |  Xiaomi | 1.0T | — | — | ||
| 18 |  Zhipu AI | 744B | 200K | $1.00 / $3.20 | ||
| 19 |  Mistral AI | 128B | 256K | $1.50 / $7.50 | ||
| 20 |  Meta | — | — | — | ||
| 21 | Alibaba Cloud / Qwen Team | 28B | 262K | $0.60 / $3.60 | ||
| 22 | Moonshot AI | 1.0T | — | — | ||
| 23 |  ByteDance | — | — | — | ||
| 24 | Alibaba Cloud / Qwen Team | 397B | 262K | $0.60 / $3.60 | ||
| 25 | OpenAI | — | 400K | $1.25 / $10.00 | ||
| 25 | OpenAI | — | 400K | $1.25 / $10.00 | ||
| 25 | OpenAI | — | — | — | ||
| 28 | Google | — | — | — | ||
| 29 | OpenAI | — | — | — | ||
| 30 | Xiaomi | — | — | — | ||
| 31 | OpenAI | — | — | — | ||
| 31 | Anthropic | — | — | — | ||
| 33 |  StepFun | 196B | 66K | $0.10 / $0.40 | ||
| 34 | Zhipu AI | 358B | — | — | ||
| 35 | OpenAI | — | — | — | ||
| 36 | ByteDance | — | — | — | ||
| 36 |  Microsoft | 1.0T | — | — | ||
| 38 | Alibaba Cloud / Qwen Team | 35B | — | — | ||
| 38 | Xiaomi | 309B | — | — | ||
| 40 | Anthropic | — | 200K | $1.00 / $5.00 | ||
| 41 | DeepSeek | 685B | — | — | ||
| 41 | DeepSeek | 685B | — | — | ||
| 41 | DeepSeek | 685B | — | — | ||
| 44 | Anthropic | — | — | — | ||
| 45 | Anthropic | — | — | — | ||
| 46 | Alibaba Cloud / Qwen Team | 27B | 262K | $0.30 / $2.40 | ||
| 47 | Alibaba Cloud / Qwen Team | 122B | 262K | $0.40 / $3.20 | ||
| 48 | Microsoft | — | — | — | ||
| 49 | Moonshot AI | 1.0T | — | — | ||
| 50 | — | 256K | $0.20 / $1.50 |
FAQ
Common questions about SWE-Bench Verified.
Sub-benchmarks
SWE-Bench Multimodal
SWE-Bench Multimodal extends SWE-Bench to evaluate language models on software engineering tasks that involve visual inputs such as screenshots, UI mockups, and diagrams alongside code understanding.
SWE-Bench Pro
SWE-Bench Pro is an advanced version of SWE-Bench that evaluates language models on complex, real-world software engineering tasks requiring extended reasoning and multi-step problem solving.
More evaluations to explore
Related benchmarks in the same category
A challenging dataset of 448 multiple-choice questions written by domain experts in biology, physics, and chemistry. Questions are Google-proof and extremely difficult, with PhD experts reaching 65% accuracy.
A more robust and challenging multi-task language understanding benchmark that extends MMLU by expanding multiple-choice options from 4 to 10, eliminating trivial questions, and focusing on reasoning-intensive tasks. Features over 12,000 curated questions across 14 domains and causes a 16-33% accuracy drop compared to original MMLU.
All 30 problems from the 2025 American Invitational Mathematics Examination (AIME I and AIME II), testing olympiad-level mathematical reasoning with integer answers from 000-999. Used as an AI benchmark to evaluate large language models' ability to solve complex mathematical problems requiring multi-step logical deductions and structured symbolic reasoning.
Massive Multitask Language Understanding benchmark testing knowledge across 57 diverse subjects including STEM, humanities, social sciences, and professional domains
Humanity's Last Exam (HLE) is a multi-modal academic benchmark with 2,500 questions across mathematics, humanities, and natural sciences, designed to test LLM capabilities at the frontier of human knowledge with unambiguous, verifiable solutions
LiveCodeBench is a holistic and contamination-free evaluation benchmark for large language models for code. It continuously collects new problems from programming contests (LeetCode, AtCoder, CodeForces) and evaluates four different scenarios: code generation, self-repair, code execution, and test output prediction. Problems are annotated with release dates to enable evaluation on unseen problems released after a model's training cutoff.