Multi-SWE Bench

A multi-language software-engineering benchmark that measures repository-level bug fixing and implementation across more than one programming ecosystem.

1 models

Higher is better.

1MiniMax M2.7MiniMax52.7

Higher is better. Bar colour: blue proprietary, orange open weights, green other.

About this benchmark

Metric: Repository task completion
Tasks: Multi-language repo tasks
Authors: MiniMax (2026)
Source: MiniMax M2.7: Early Echoes of Self-Evolution ↗

Scores are as reported by model developers or evaluators and may use different settings (for example reasoning effort or tools). Compare with care.