One API. The right model for each request.

Choose models by task

Route a short classification task to a model that meets your cost and latency target. Choose another for complex analysis or code generation. Store these choices in server-side configuration so you can update them without changing the user interface.

Include GPT, Claude and Gemini in the same evaluation as Kimi, GLM, DeepSeek and MiMo. The useful comparison is the result on your workload, not whether a model has open weights.

Keep routing rules explicit

Use the model field to select a supported identifier for each call. Make a small mapping from application tasks to model identifiers, then log which model served each request.

A common endpoint does not require identical settings for every model. Keep model-specific capabilities and reasoning parameters with the routing configuration. Avoid forwarding unvalidated model names from public clients.

Keep regional rules in the routing decision

For a workload that requires EU inference, select the EU-hosted routes enabled for it. Restrict the eligible models before comparing performance. A fallback should remain within the same processing requirements.

Use the data residency guide to document the route and the systems around it. The general model catalog and the EU-eligible list serve different purposes.

Make your first API request

Create a TextCortex account and generate an API key in account settings. Install the OpenAI Python package on your server. Store the key in TEXTCORTEX_API_KEY and set TEXTCORTEX_MODEL to an identifier returned by GET /models.

This example keeps credentials in environment variables. The same client configuration works when you choose another supported model. Read the API reference for endpoint-specific parameters.

import os
from openai import OpenAI

client = OpenAI(
    api_key=os.environ["TEXTCORTEX_API_KEY"],
    base_url="https://api.textcortex.com/v1",
)

for model in client.models.list():
    print(model.id)

response = client.chat.completions.create(
    model=os.environ["TEXTCORTEX_MODEL"],
    messages=[{"role": "user", "content": "Hello"}],
)
print(response.choices[0].message.content)