FrontierThe story, in brief

Xiaomi's affordable flagship AI leads the open models, and Anthropic says Claude helped get it there

Xiaomi's MiMo-V2.6-Pro tops open models at $2.62M training cost. Anthropic says Claude's training data got siphoned in the process.

Illustration of a transparent lens revealing connected networks across layers of paper.
Exploring the next frontier of AI research.AI illustration by KeyNews
The KeyNews take

Why it matters

A new open-weight model has broken into the frontier-capability tier at a fraction of expected cost, but the achievement rests on contested training practices — raising questions about data sourcing and whether open models can compete on performance without relying on proprietary model distillation.

The key facts

5 to know
  1. MiMo-V2.6-Pro ranks first among openly available AI models

  2. Training cost: $2.62 million via reinforcement learning

  3. Significant price undercut vs. competitors

  4. Anthropic alleges Claude training data was used without permission

  5. Lab-race drama: capability claims shadowed by training-data accusations

Go to the source

The Decoderthe-decoder.com

Publisher excerpt: With MiMo-V2.6-Pro, Xiaomi moves to the top of the openly available AI models and drastically undercuts the competition on price. What makes that possible is massive reinforcement learning that cost $2.62 million. But the success comes with a catch: Anthropic accuses the company of siphoning off…
Read original report
Back to today's editionMore frontier news

The wider picture

View all
Illustration of a transparent lens revealing connected networks across layers of paper.
AI illustration by KeyNews
Frontier01

OpenAI launches GPT-6 Sol and Luna, boasting lower cost and fewer mistakes

Two-model strategy signals OpenAI's bet on specialization over one-size-fits-all frontier capability. Practitioners choosing between cost and quality now have official paths; enthusiasts watch if this reshapes the lab-race playbook.

TechCrunch AI
Illustration of a transparent lens revealing connected networks across layers of paper.
AI illustration by KeyNews
Frontier02

Claude Opus 5.5 matches Fable 5.1 performance at lower cost and promises less "Claudish" writing

A new generation of Claude models arrives with meaningful cost reduction and claimed capability parity to Anthropic's previous flagship, while positioning against OpenAI's latest. This matters for practitioners choosing between models and for understanding the efficiency frontier in the lab race.

The Decoder
Illustration of a transparent lens revealing connected networks across layers of paper.
AI illustration by KeyNews
Frontier03

Anthropic releases Opus 5.5 with lower prices and Fable-level performance

A new flagship model from a frontier lab claims best-in-class performance while undercutting rivals on price—a capability + economics shift that reshapes the competitive landscape and forces practitioners to re-evaluate their model strategies.

TechCrunch AI