Runware Adds MiniMax M2.7 on Day Zero of Release

Mar 20, 2026 · Updated Apr 25, 2026

Runware has added MiniMax M2.7 — MiniMax's long-context LLM designed for agentic workflows across software engineering, tool use, and office productivity — to its model catalog on day zero of release. The model features a self-evolving architecture that delivers 30% performance gains across 100+ iterations and a 97% skill adherence rate across 40+ complex tasks. Benchmarks include 56.22% on SWE-Pro, 55.6% on VIBE-Pro, and 57.0% on Terminal Bench 2. A Highspeed variant is also available.

Runware lists M2.7 at $0.30/M input and $1.20/M output tokens with commercial use permitted. The Highspeed variant keeps the same output behavior while reducing latency for interactive coding agents, tool-calling pipelines, and office automation flows.

Developers building agentic coding or document automation workflows can start running M2.7 through Runware's API today, choosing between the standard and Highspeed variants based on latency requirements.

View the full update on runware.ai

Runware

@runwareMar 18

MiniMax M2.7 is live on Runware on Day0! 🚀 - full end-to-end project delivery - self-evolving architecture: 30% performance gains across 100+ iterations - 97% skill adherence rate across 40+ complex tasks - 56.22% on SWE-Pro, 55.6% on VIBE-Pro, 57.0% on Terminal Bench 2 https://t.co/g1QW0C8G00

View on X

Keep reading

Keep reading

MiniMax Launches M2.7, Its First Self-Evolving Agentic Model

MiniMax-M2.7 Now Available on Ollama Cloud for Agentic Coding

MiniMax M2.7 Enters Code Arena Top 10 as Most Cost-Efficient Model

ASI:Cloud Adds MiniMax M2.5, Qwen, and GLM Models for Inference

Keep reading

MiniMax Launches M2.7, Its First Self-Evolving Agentic Model

MiniMax-M2.7 Now Available on Ollama Cloud for Agentic Coding

MiniMax M2.7 Enters Code Arena Top 10 as Most Cost-Efficient Model

ASI:Cloud Adds MiniMax M2.5, Qwen, and GLM Models for Inference