Skip to main content

Documentation Index

Fetch the complete documentation index at: https://doc.hitopen.com/llms.txt

Use this file to discover all available pages before exploring further.

Newapi does not have a fixed model catalog. The models available to you depend entirely on which channels an administrator has configured. When a channel is added and assigned a list of model names, those names become callable through Newapi’s standard API endpoints.

​
How model availability works

When you call POST /v1/chat/completions with a model name, Newapi looks up which channels list that model, selects one based on priority and weight, and forwards your request. From your application’s perspective, you are calling a standard OpenAI-compatible endpoint — the upstream routing is transparent. This design means:
  • Model names are administrator-defined. Your admin decides which names are exposed and which upstream provider each name maps to.
  • The same model name can be served by multiple providers. Newapi load-balances across channels automatically.
  • Custom aliases are possible. An admin might configure fast-chat to route to a specific model on a specific provider.
If a model you need is not available, contact your Newapi administrator to have the corresponding channel configured.

​
Listing available models

Use the standard OpenAI GET /v1/models endpoint to retrieve the models your token can access. 
curl "https://YOUR_NEWAPI_BASE_URL/v1/models" \
  -H "Authorization: Bearer YOUR_API_KEY"
The response follows the standard OpenAI format: 
{
  "object": "list",
  "data": [
    {
      "id": "gpt-4o",
      "object": "model",
      "created": 1715000000,
      "owned_by": "openai"
    },
    {
      "id": "gemini-2.0-flash",
      "object": "model",
      "created": 1715000000,
      "owned_by": "google"
    }
  ]
}
The id field is the model name you pass in API requests. Use it exactly as shown — model names are case-sensitive.

​
Model names and routing

Model names in Newapi work as follows:
  • Standard names (e.g., gpt-4o, gemini-2.5-pro) map directly to the corresponding upstream model on the configured channel.
  • Custom aliases (e.g., fast-chat, my-org-default) are defined by your administrator and can point to any model on any provider. Check with your admin if you see unfamiliar model names.
  • Model names are scoped to your token. If your token has model restrictions, GET /v1/models only returns models that token is allowed to use.
Always retrieve the model list from /v1/models rather than hardcoding model names. This ensures your application works with whatever models are currently available on your Newapi instance.

​
Checking model pricing

To see the token pricing for each model, call the /api/pricing endpoint: 
curl "https://YOUR_NEWAPI_BASE_URL/api/pricing" \
  -H "Authorization: Bearer YOUR_API_KEY"
The response includes input and output token costs per model. Use this to estimate costs before running large workloads.

​
Next steps

  • Channels — Understand how channels control which models are available
  • Usage Logs — Track model usage and costs per request
  • API Reference: Models — Full endpoint reference for GET /v1/models