Kimi K2.6

Kimi K2.6 is a native multimodal agentic system built on a 1 trillion parameter mixture-of-experts architecture with 32 billion active parameters — 384 experts, 8 routed plus 1 shared per token — and a MoonViT vision encoder that handles image and video in the same stack rather than through a separate module. It excels at long-horizon coding, multi-step reasoning, and tool use, and its Agent Swarm system scales to 300 domain-specialized sub-agents executing up to 4,000 coordinated steps in a single autonomous run.

Features

On-demand Deployments

On-demand deployments let you run Kimi K2.6 on dedicated GPUs with Sciforium's high-performance serving stack, with high reliability and no rate limits.

Docs

Agent Swarm Orchestration

Coordinates up to 300 domain-specialized sub-agents across 4,000 steps in one autonomous run — triple the swarm size of K2.5.

Docs
MiniMax  M2.5
Kimi K2.5
GLM 5
DeepSeek V3.2
gpt-oss-120b
gpt-oss-20b
Qwen3 Instruct
Qwen3 Thinking
Qwen3 Coder
Qwen3.5
Qwen3 VL Instruct
Qwen3 ASR
Qwen-Image
Qwen-Image-Edit
Flux2
Stable Diffusion 3.5
Hunyuan Image
Z-Image
Wan2.2-I2V
Wan2.2-T2V
Hunyuan Image
Z-Image