SWE-MM

Name: SWE-MM Leaderboard — AI Model Scores
Creator: LLM Stats
License: https://llm-stats.com/legal/terms-of-service

Paper Implementation

Progress Over Time

Interactive timeline showing model performance evolution on SWE-MM

State-of-the-art frontier

Open

Proprietary

SWE-MM Leaderboard

1 models

				Context	Cost	License
1	Qwen3.8-27B Alibaba Cloud / Qwen Team		28B	—	—

Notice missing or incorrect data?

About this benchmark

What is SWE-MM?

SWE-MM evaluates software-engineering agents on repository tasks that require understanding both source code and visual evidence.

SWE-MM is a multimodal benchmark evaluating models on multimodal, agents, code, and vision tasks. LLM Stats tracks 1 models on this benchmark, scored on a 0–1 scale. The current average is 0.4, with the leader at 0.4.

Compare leaders on the best AI for multimodal, best AI for agents, best AI for code and best AI for vision leaderboards.

Current leaders

Qwen3.8-27B from Alibaba Cloud / Qwen Team currently leads the SWE-MM leaderboard with a score of 0.386 across 1 evaluated AI models.

Qwen3.8-27BAlibaba Cloud / Qwen Team38.6%

FAQ

Common questions about the SWE-MM benchmark and leaderboard.

What is the SWE-MM benchmark?

SWE-MM evaluates software-engineering agents on repository tasks that require understanding both source code and visual evidence.

What is the SWE-MM leaderboard?

The SWE-MM leaderboard ranks 1 AI models based on their performance on this benchmark. Currently, Qwen3.8-27B by Alibaba Cloud / Qwen Team leads with a score of 0.386. The average score across all models is 0.386.

What is the highest SWE-MM score?

The highest SWE-MM score is 0.386, achieved by Qwen3.8-27B from Alibaba Cloud / Qwen Team.

How many models are evaluated on SWE-MM?

1 models have been evaluated on the SWE-MM benchmark, with 0 verified results and 1 self-reported results.

Where can I find the SWE-MM paper?

The SWE-MM paper is available at https://www.swebench.com/multimodal.html. The paper details the methodology, dataset construction, and evaluation criteria.

Where can I find the SWE-MM dataset?

The SWE-MM dataset is available at https://github.com/SWE-bench/SWE-bench.

What categories does SWE-MM cover?

SWE-MM is categorized under multimodal, agents, code, and vision. The benchmark evaluates multimodal models.

What is the best open-source model on SWE-MM?

Qwen3.8-27B by Alibaba Cloud / Qwen Team is the top-ranked open-source model on SWE-MM, with a score of 0.386 (rank #1).

How recent are the SWE-MM leaderboard results?

The SWE-MM leaderboard was last updated in August 2026 and currently includes 1 evaluated models.