Model Platform
Image & Video Models
Create and scale visual imagination.
Large Language Models
Power intelligent reasoning and conversation.
Serverless
Deploy, customize, and build with Dedicated Endpoint, Fine-Tuning, and DevPod.
Featured Model Families
View All Models
OpenAI
Anthropic
Deepseek
Google
Qwen & Wan
Bytedance
Kling
Minimax
Agent Solutions
GPU Cloud
On-Demand GPUs
Scale AI workloads with high-performance cloud GPUs.
Bare Metal
Run AI models faster on dedicated GPU servers.
Pricing
Subscription Plan
Access simple monthly plans for developers.
Pay by Usage
Pay only for what you generate.
Pay by GPU/hr
Scale AI with on-demand GPU compute.
Developers
Docs
Explore guides, APIs, and examples to build on Atlas Cloud.
Blog
Stay updated with product releases, insights, and tutorials.
Company
About Us
Discover who we are and why we build.
Our Team
Meet the talent building Atlas Cloud.
News & Events
Explore our latest updates and announcements.
Contact
Contact
Model Platform
Image & Video Models
Create and scale visual imagination.
Large Language Models
Power intelligent reasoning and conversation.
Serverless
Deploy, customize, and build with Dedicated Endpoint, Fine-Tuning, and DevPod.
Agent Solutions
GPU Cloud
On-Demand GPUs
Scale AI workloads with high-performance cloud GPUs.
Bare Metal
Run AI models faster on dedicated GPU servers.
Pricing
Subscription Plan
Access simple monthly plans for developers.
Pay by Usage
Pay only for what you generate.
Pay by GPU/hr
Scale AI with on-demand GPU compute.
Developers
Docs
Explore guides, APIs, and examples to build on Atlas Cloud.
Blog
Stay updated with product releases, insights, and tutorials.
Company
About Us
Discover who we are and why we build.
Our Team
Meet the talent building Atlas Cloud.
News & Events
Explore our latest updates and announcements.
moonshotai/Kimi-K2-Instruct-0905
LLM
HOT
NEW
Model Platform
moonshotai/Kimi-K2-Instruct-0905
moonshotai/
Kimi-K2-Instruct-0905
Loading...
All You Need to Know About This Model
Overview:
Model Provider:
MOONSHOTAI
Model Type:
Large Language Model
Deployment:
Inferencing API; Playground
Pricing:
$0.6/M Input & $2.5/M Output
Key Specs:
Parameters:
1.0T total; ~32B active
Context:
65k tokens
Architecture Type:
Mixture-of-Experts (MoE)
Knowledge Cutoff:
April 2025
Core Strengths:
Enhanced agentic coding (53.7% LiveCodeBench)
Improved frontend development capabilities
Extended 256K context window
Excellent tool calling abilities
Use Cases:
Advanced frontend development
Sophisticated AI agents
Full-stack development
Complex long-context workflows
Start From 300+ Models,
Only at Atlas Cloud.
Explore all models