Kimi K2 0905 is an updated version of Kimi K2, a state-of-the-art mixture-of-experts (MoE) language model with 32 billion activated parameters and 1 trillion total parameters. Kimi K2 0905 has improved coding abilities, a longer context window, and agentic tool use, and a longer (262K) context window.
Fine-tuningDocs | Kimi K2 Instruct 0905 can be customized with your data to improve responses. Fireworks uses LoRA to efficiently train and deploy your personalized model |
On-demand DeploymentDocs | On-demand deployments allow you to use Kimi K2 Instruct 0905 on dedicated GPUs with Fireworks' high-performance serving stack with high reliability and no rate limits. |
Kimi K2 Instruct 0905 is a state-of-the-art mixture-of-experts (MoE) language model developed by Moonshot AI. It features 32 billion activated parameters out of 1 trillion total parameters, offering enhanced performance for agentic tasks and long-context reasoning.
Kimi K2 Instruct 0905 is optimized for:
The model supports a maximum context length of 262,144 tokens on Fireworks.
The model's usable context window is 256,000 tokens.
Yes. Multiple quantized versions are available.
The default temperature of Kimi K2 Instruct 0905 is 0.6.
Yes. Fireworks supports fine-tuning Kimi K2 Instruct 0905 using LoRA for parameter-efficient adaptation.
Kimi K2 Instruct 0905 is released under a Modified MIT License.
