Same contract as every other model
Public API IDs, Bearer keys, streaming SSE — no Moonshot-only SDK required.
Kimi K3, K2.7 Code, and K2.6 over the same OpenAI-compatible endpoint you already use for DeepSeek and Qwen.
Why SupaNexus
Public API IDs, Bearer keys, streaming SSE — no Moonshot-only SDK required.
K3 for flagship agent work, K2.7 Code for repositories, K2.6 when you need multimodal context.
Cap spend per project. Prompts are not retained for training.
Catalog
Send the Public API ID as model. Prefix is moonshot/ even when you search for Kimi.
| Model | Public API ID | Context | Best for |
|---|---|---|---|
| Kimi K3 | moonshot/kimi-k3 | 1M | FlagshipMultimodalAgents |
| Kimi K2.7 Code | moonshot/kimi-k2.7-code | 256K | CodingThinking |
| Kimi K2.6 | moonshot/kimi-k2.6 | 256K | MultimodalAgents |
Live rates are in the developer console. See also homepage pricing · console catalog
Quickstart
Official OpenAI SDK. Only base URL, API key, and model ID change.
https://api.supanexus.ai/v1
Python
from openai import OpenAI
client = OpenAI(
base_url="https://api.supanexus.ai/v1",
api_key="sk-snx-...",
)
response = client.chat.completions.create(
model="moonshot/kimi-k3",
messages=[{"role": "user", "content": "Hello!"}],
)
print(response.choices[0].message.content)
FAQ
IDs, compatibility, and how Kimi sits next to other vendors.
Kimi models are published under the Moonshot vendor in the SupaNexus catalog. Always send the full Public API ID, including the moonshot/ prefix.
Use moonshot/kimi-k3 as the default flagship. Pick moonshot/kimi-k2.7-code for repository and coding agents.
Yes. Point it at https://api.supanexus.ai/v1 with a SupaNexus project key.
Create a SupaNexus project, copy the key, and send moonshot/kimi-k3 on Chat Completions.