MiniMax Just Open Sourced MiniMax M2.7: A Self-Evolving Agent Model that Scores 56.22% on SWE-Pro and 57.0% on Terminal Bench 2

MarkTechPost / 4/12/2026

📰 NewsSignals & Early TrendsIndustry & Market MovesModels & Research

Key Points

  • MiniMax has open-sourced its MiniMax M2.7 model by publishing the model weights on Hugging Face, marking a new release for public use and community experimentation.
  • The company describes M2.7 as its most capable open-source model to date and its first agent model that can actively participate in its own development cycle.
  • MiniMax M2.7 is reported to achieve 56.22% on SWE-Pro and 57.0% on Terminal Bench 2, indicating strong performance on software engineering and terminal-based evaluation tasks.
  • The release represents a shift toward “self-evolving” agent development workflows, which could influence how future LLM/agent systems are trained and iterated.

MiniMax has officially open-sourced MiniMax M2.7, making the model weights publicly available on Hugging Face. Originally announced on March 18, 2026, MiniMax M2.7 is the MiniMax’s most capable open-source model to date — and its first model to actively participate in its own development cycle, a meaningful shift in how large language models are built […]

The post MiniMax Just Open Sourced MiniMax M2.7: A Self-Evolving Agent Model that Scores 56.22% on SWE-Pro and 57.0% on Terminal Bench 2 appeared first on MarkTechPost.