Inferway

Inferway is an inference API for open-weight models, running on our own GPUs. It speaks both the OpenAI and the Anthropic formats.

Models

• XiaomiMiMo/MiMo-V2.6-Flash-RL as inferway/mimo-v2.6-flash: chat with streaming, tool calling and JSON-schema structured outputs
• MiniMaxAI/MiniMax-H3 as inferway/minimax-h3-768p: 768P video generation with audio, billed per video second
• More open-source models are on the way

API

• OpenAI format: POST https://api.inferway.ai/v1/chat/completions
• Anthropic format: POST https://api.inferway.ai/v1/messages

Links: pricing · docs · hello@inferway.ai

# OpenAI SDK
from openai import OpenAI

client = OpenAI(base_url="https://api.inferway.ai/v1", api_key="YOUR_INFERWAY_API_KEY")
resp = client.chat.completions.create(
    model="inferway/mimo-v2.6-flash",
    messages=[{"role": "user", "content": "Hello!"}],
)
print(resp.choices[0].message.content)

# Anthropic SDK
from anthropic import Anthropic

client = Anthropic(base_url="https://api.inferway.ai", api_key="YOUR_INFERWAY_API_KEY")
msg = client.messages.create(
    model="inferway/mimo-v2.6-flash",
    max_tokens=1024,
    messages=[{"role": "user", "content": "Hello!"}],
)
print(msg.content[0].text)