FlashNEWS BRIEFING

Xiaomi Open-Sources Terminal AI Coding Assistant MiMo Code

1 min read

88Z AI, 11 June — Xiaomi's large model team has launched MiMo Code, an open-source terminal AI coding assistant. It features the time-limited free MiMo-V2.5 model and supports mainstream models like DeepSeek and Kimi. Key capabilities include a persistent memory system, Compose mode for full-process automated development, memory organization commands, and voice control. In SWE-Bench Pro and Terminal Bench 2 benchmarks, it outperforms Claude Code under identical model conditions.

At a glance

  • Bundled with MiMo-V2.5 model and supports mainstream models like DeepSeek and Kimi
  • Outperforms Claude Code on SWE-Bench Pro and Terminal Bench 2 benchmarks
Cover image for Xiaomi Open-Sources Terminal AI Coding Assistant MiMo Code
Cover image for Xiaomi Open-Sources Terminal AI Coding Assistant MiMo Code

The update

Xiaomi's large model team has officially open-sourced MiMo Code, a terminal-based AI coding assistant. The tool comes with the MiMo-V2.5 model, available for free for a limited time, and is compatible with mainstream industry models such as DeepSeek and Kimi, offering users flexible options. The core highlights of MiMo Code include its proprietary persistent memory system, which retains context over the long term. Its Compose mode enables fully automated development from requirements to code. Additionally, users can organize their memory banks periodically using the /dream command and control the assistant via voice input. Benchmark data shows that on SWE-Bench Pro and Terminal Bench 2, MiMo Code outperforms Claude Code when using the same underlying model.

Sources & editorial note

This article is an AI-assisted synthesis by 88Z AI based on the public sources below. Product claims and availability are attributed to their publishers. It does not represent independent testing or first-hand reporting.

  1. Xiaomi MiMoSource published

Source dates above refer to the original announcements; the date under the headline is this article’s publication on 88Z AI.

Related reading

All stories
88Z.ai