How to call any catalog model through the API

One API, any catalog id

Developers calling models through AshnaAI use an OpenAI-compatible HTTP API. Base URL: https://api.ashna.ai/v1/api Auth is Authorization: Bearer YOUR_API_KEY. The model field is either a foundation catalog id or a custom agent id. That is how GLM-5.3-Flash, GPT-5.6 Terra, a Claude row, Kimi K3, or Ashna-X1 are called from code—one route, not a URL per vendor.

Official reference: AshnaAI API docs, including model and agent ids. If you do not have a key, create one in Account → API.

Step 1: Get a key

Open Account → API and choose Create key. Send it as a Bearer token. Do not put the key in a public blog, client bundle, or screenshot.

Step 2: List models, then pick an id

GET https://api.ashna.ai/v1/api/models returns the OpenAI list shape. Copy the id string exactly as listed—for Flash that is glm-5.3-flash. The same id is visible in the in-app model picker. Custom agents use the agent id on the same chat-completions path. Endpoint notes: list models.

Step 3: POST chat completions

POST https://api.ashna.ai/v1/api/chat/completions with model, messages, and optional stream, temperature, or max_tokens. To switch from Flash to GPT-5.6 Terra, change model to gpt-5.6-terra. To run a saved agent, pass that agent’s customAgentId as model. Endpoint notes: chat completions.

To try Flash in the product UI instead of curl, open GLM-5.3-Flash in chat. For when to pin Flash versus other rows, start at Pin GLM-5.3-Flash for cost-effective coding.

Try the API now

Create a key in Account → API, then open AshnaAI API docs and run the chat-completions examples with the model id you listed. To use the same key from the editor, follow How to use the AshnaAI API in VS Code.