Models¶

The models endpoints provide catalogs of available LLM and embedding models from OpenRouter, OpenAI, and Ollama. These are used by the frontend for model selection dropdowns.

Base path: /api/models

GET /api/models/openrouter¶

Fetch available LLM models from the OpenRouter API. Falls back to a curated list of 23 popular models when the API is unavailable or no API key is configured.

Auth: User

RequestResponseErrorscurl ExamplePython Example

Headers:

Header	Value
`Authorization`	`Bearer <token>`

Status: 200 OK

{
  "success": true,
  "data": {
    "models": [
      {
        "id": "openai/gpt-4o-mini",
        "name": "GPT-4o Mini",
        "provider": "OpenAI",
        "context": 128000,
        "pricing": {
          "prompt": "0.00000015",
          "completion": "0.0000006"
        }
      },
      {
        "id": "anthropic/claude-sonnet-4",
        "name": "Claude Sonnet 4",
        "provider": "Anthropic",
        "context": 200000
      }
    ],
    "source": "api",
    "total": 150
  }
}

Field	Type	Description
`source`	string	`"api"` (live OpenRouter), `"curated"` (no API key), or `"curated_fallback"` (API error)
`total`	integer	Number of models (present when `source` is `"api"`)

Code	Cause
`401`	Invalid or missing token

curl http://localhost:8000/api/models/openrouter \
  -H "Authorization: Bearer $TOKEN"

import httpx

response = httpx.get(
    "http://localhost:8000/api/models/openrouter",
    headers={"Authorization": f"Bearer {token}"},
)
data = response.json()["data"]
print(f"Source: {data['source']}, Models: {len(data['models'])}")
for m in data["models"][:5]:
    print(f"  {m['id']} ({m['provider']}, {m['context']} ctx)")

GET /api/models/openrouter/embeddings¶

Fetch available embedding models from the OpenRouter API. Falls back to a curated list of 21 models from 8 providers.

Auth: User

RequestResponseErrorscurl ExamplePython Example

Headers:

Header	Value
`Authorization`	`Bearer <token>`

Status: 200 OK

{
  "success": true,
  "data": {
    "models": [
      {
        "id": "openai/text-embedding-3-small",
        "name": "OpenAI Text Embedding 3 Small",
        "provider": "OpenAI",
        "dimensions": 1536,
        "context": 8192,
        "cost_per_million": "$0.02"
      },
      {
        "id": "baai/bge-m3",
        "name": "BGE-M3 (Multilingual)",
        "provider": "BAAI",
        "dimensions": 1024,
        "context": 8192,
        "cost_per_million": "$0.01"
      }
    ],
    "source": "curated"
  }
}

Field	Type	Description
`source`	string	`"api"`, `"curated"`, or `"curated_fallback"`

Embedding Model Immutability

The embedding model is selected when creating a notebook and cannot be changed. This endpoint is used to populate the model dropdown in the Create Notebook dialog.

Code	Cause
`401`	Invalid or missing token

curl http://localhost:8000/api/models/openrouter/embeddings \
  -H "Authorization: Bearer $TOKEN"

import httpx

response = httpx.get(
    "http://localhost:8000/api/models/openrouter/embeddings",
    headers={"Authorization": f"Bearer {token}"},
)
data = response.json()["data"]
for m in data["models"]:
    print(f"  {m['id']}: {m['dimensions']}d, {m['cost_per_million']}/M tokens")

GET /api/models/ollama¶

Fetch locally-pulled models from the Ollama instance. Filters out embedding-only models (nomic-embed-text, mxbai-embed, etc.) and returns only inference models.

Auth: User

RequestResponseErrorscurl ExamplePython Example

Headers:

Header	Value
`Authorization`	`Bearer <token>`

Status: 200 OK

{
  "success": true,
  "data": {
    "models": [
      {
        "id": "qwen2.5:1.5b",
        "name": "qwen2.5:1.5b",
        "parameter_size": "1.5B",
        "quantization": "Q4_K_M",
        "family": "qwen2",
        "size_mb": 986
      }
    ],
    "source": "ollama",
    "total": 1
  }
}

Field	Type	Description
`source`	string	`"ollama"` (connected) or `"ollama_unreachable"` (cannot connect)
`error`	string	Error message when Ollama is unreachable (only present on error)

Code	Cause
`401`	Invalid or missing token

curl http://localhost:8000/api/models/ollama \
  -H "Authorization: Bearer $TOKEN"

import httpx

response = httpx.get(
    "http://localhost:8000/api/models/ollama",
    headers={"Authorization": f"Bearer {token}"},
)
data = response.json()["data"]
if data["source"] == "ollama_unreachable":
    print(f"Ollama not running: {data.get('error')}")
else:
    for m in data["models"]:
        print(f"  {m['name']} ({m['parameter_size']}, {m['size_mb']}MB)")