2 min read

LM Studio Bionic adds Moonshot’s Kimi K3

LM Studio Bionic now supports Moonshot AI’s Kimi K3, a 2.8-trillion-parameter model with a 1-million-token context window.

Image: 9to5Mac

LM Studio’s Bionic now supports Kimi K3, Moonshot AI’s open model with 2.8 trillion parameters and a 1-million-token context window. The update gives users access to a larger model for coding, research, and complex tasks involving documents and files.

Kimi K3 comes to Bionic

LM Studio introduced Bionic a few days ago as an agentic workflow app for running open models. Users can work with models locally or switch to cloud-hosted options when they need longer context windows, stronger reasoning, or more capable tool calling.

The company says Bionic is designed for “getting real work done with open models,” while allowing users to control their privacy and AI spending. Its cloud models use US-based inference with Zero Data Retention (ZDR), meaning prompts, files, and outputs are not stored after processing.

Before Kimi K3, Bionic supported DeepSeek V4, Pro, GLM-5.2, Kimi K2.6, and Kimi-K2.7-Code. LM Studio announced that Kimi K3 was live on July 27, 2026:

Recommended reading

Anthropic rejects calls to ban open-weight AI

“A massive, open frontier model: 2.8T MoE with 1M context window. Kimi K3 in Bionic is hosted on US-based servers, with ZDR enabled by default.”

LM Studio, July 27, 2026

Kimi K3 pricing through Bionic is:

  • $3 per million input tokens
  • $0.30 per million cached input tokens
  • $15 per million output tokens

That is higher than the previously available models, which ranged from $0.95 to $.174 per million input tokens and $3.48 to $4.50 per million output tokens, according to the source.

Bionic is available for Mac and Windows.

Ava Chen

AI Editor

Ava covers the rapidly evolving world of artificial intelligence, from foundational models and research labs to the real-world economics of intelligence. With a background in computational linguistics, she cuts through the hype to find out what actually works. She firmly believes that benchmarks are just marketing until reproduced in the wild.

via 9to5Mac

/ Keep reading