Xiaomi's open-weight MiMo-V2.5-Pro takes aim at Claude Opus with hours-long autonomous coding

THE DECODER / 5/3/2026

📰 NewsIndustry & Market MovesModels & Research

共有:

Key Points

Xiaomi has released the open-weight MiMo-V2.5-Pro, claiming it nearly matches Anthropic’s Claude Opus 4.6 on coding benchmarks.
Xiaomi says MiMo-V2.5-Pro uses 40–60% fewer tokens, suggesting improved efficiency for coding tasks.
The company positions the model’s “hours-long autonomous coding” as a key differentiator, emphasizing long-duration agentic performance rather than only scores.
The release reflects a broader shift among Chinese open-weight providers, where competitive advantage is moving toward cost and autonomy/runtime on single tasks.
The push puts additional pressure on the open-weight landscape led by players such as Deepseek, intensifying competition on both capability and operating expense.

Xiaomi MiMo intro foil with large "Hello, I'm MiMo" lettering against a background with a repeated "MIMO" pattern.

Xiaomi's new MiMo-V2.5-Pro nearly matches Anthropic's Claude Opus 4.6 on coding benchmarks while burning 40 to 60 percent fewer tokens, according to the company. The release pushes Xiaomi deeper into the race among Chinese open-weight providers like Deepseek, where the fight is shifting from raw benchmark scores to how cheaply and how long a model can run autonomously on a single task.

The article Xiaomi's open-weight MiMo-V2.5-Pro takes aim at Claude Opus with hours-long autonomous coding appeared first on The Decoder.