Quickstart
Typno AI runs open-weight language and image models in EU datacenters. Use it from the chat, or from your own code through an API that speaks the OpenAI and Anthropic formats.
1. Get a key
Order an AI API service in My Typno and create a key under API keys. Keys start with tpk-. A service has no fixed costs; you pay per token.
2. Send a request
curl https://dev.typno.ai/v1/chat/completions \
-H "Authorization: Bearer $TYPNO_API_KEY" \
-H "Content-Type: application/json" \
-d '{"model": "mistral4-119b", "messages": [{"role": "user", "content": "Hello!"}]}'
With the OpenAI SDK only the address changes:
from openai import OpenAI
client = OpenAI(base_url="https://dev.typno.ai/v1", api_key="tpk-...")
reply = client.chat.completions.create(
model="mistral4-119b",
messages=[{"role": "user", "content": "Hello!"}],
)
print(reply.choices[0].message.content)
Addresses
| What | Address |
|---|---|
| OpenAI-style API | https://dev.typno.ai/v1 |
| Anthropic-style API | https://dev.typno.ai (the SDK adds /v1/messages) |
| MCP server | https://dev.typno.ai/mcp |
| Chat | https://dev.typno.ai/chat |
Next
- Models and pricing
- Connect an agent: Claude Code, Codex, opencode, Hermes
- Where your data stays