Semifly
Open-weight · Moonshot AI

Kimi K2.7-Code

Moonshot's agentic, coding-focused 1-trillion-parameter MoE (32B active) with a 256K context window — the K2.7 update (Jun 12) cuts reasoning token usage ~30% versus K2.6 and boosts MCP tool-calling, excelling at long documents and long-horizon coding.

Developer
Moonshot AI
Parameters
~1T (A32B, MoE)
Context
256K
License
Mod. MIT
Run on Semifly →Download on Hugging Face

About Kimi K2.7-Code

Moonshot's agentic, coding-focused 1-trillion-parameter MoE (32B active) with a 256K context window — the K2.7 update (Jun 12) cuts reasoning token usage ~30% versus K2.6 and boosts MCP tool-calling, excelling at long documents and long-horizon coding.

Parameters, context, and license as reported July 2026. Confirm the current license on the official Hugging Face page before deployment.

← Back to all models

Run any of these on Semifly

Tokens & API

Access hosted models through a simple, metered token API.

Get API access →

GPU servers

Buy or lease Supermicro GPU systems to self-host open-weight models.

Browse GPU servers →

AI Foundry

Managed compute for training, fine-tuning, and inference.

Explore AI Foundry →