Use a catalog that spans model families
Frontier model access is not limited to a single licensing or distribution model. TextCortex connects your application to GPT, Claude and Gemini alongside Kimi, GLM, DeepSeek, MiMo and other supported families.
Retrieve available identifiers from GET /models and select the version you intend to use. Keep availability and capability checks in the release process as the catalog evolves.
Select European deployments for European requirements
Use the EU-hosted route for a workload that requires regional inference. Keep that choice explicit when evaluating a newer version or introducing a second model.
Explore the EU model pages for the dedicated Kimi K3, GLM-5.3, DeepSeek and MiMo integrations. Other catalog routes can have different processing arrangements.
Upgrade a model without redesigning the client
Use a common TextCortex API client with the model identifier stored in configuration. Introduce a candidate model in a controlled evaluation while the existing route remains available.
Compare the outputs and request capabilities your application actually depends on. A newer version can change costs, latency or generation behavior even when the API shape stays familiar.
Build the first request, then the evaluation
Start with the OpenAI-compatible quickstart and a TextCortex API key. Add task-specific instructions and output validation once the basic connection works.
Use the model selection guide to compare candidates and the production rollout guide to introduce the selected model.