Skip to main content
GET
/
api
/
v1
/
ollama
/
model-details
curl --request GET \
  --url 'http://localhost:3000/api/v1/ollama/model-details?name=llama2&verbose=true' \
  --header 'Authorization: Bearer <token>'

{
  "success": true,
  "model": {
    "name": "llama2",
    "summary": "Meta's Llama 2 model for general chat and text generation",
    "downloads": "5.4M",
    "tags": ["latest", "7b", "13b", "70b"],
    "lastUpdated": "2024-01-10T15:30:00Z",
    "sizes": ["3.8GB", "7.4GB", "40GB"],
    "variants": [
      {
        "name": "llama2:7b",
        "size": "3.8GB",
        "parameters": "7B"
      },
      {
        "name": "llama2:13b",
        "size": "7.4GB",
        "parameters": "13B"
      }
    ],
    "readme": "# Llama 2\n\nMeta's Llama 2...",
    "command": "ollama run llama2",
    "url": "https://ollama.com/library/llama2"
  }
}

​
Query Parameters

​
name
string
required
Model name to get details for (e.g., “llama2”, “mistral”)
​
verbose
boolean
default:false
Include verbose information in the response

​
Response

​
success
boolean
Indicates if the request was successful
​
model
object
Detailed model information
curl --request GET \
  --url 'http://localhost:3000/api/v1/ollama/model-details?name=llama2&verbose=true' \
  --header 'Authorization: Bearer <token>'

{
  "success": true,
  "model": {
    "name": "llama2",
    "summary": "Meta's Llama 2 model for general chat and text generation",
    "downloads": "5.4M",
    "tags": ["latest", "7b", "13b", "70b"],
    "lastUpdated": "2024-01-10T15:30:00Z",
    "sizes": ["3.8GB", "7.4GB", "40GB"],
    "variants": [
      {
        "name": "llama2:7b",
        "size": "3.8GB",
        "parameters": "7B"
      },
      {
        "name": "llama2:13b",
        "size": "7.4GB",
        "parameters": "13B"
      }
    ],
    "readme": "# Llama 2\n\nMeta's Llama 2...",
    "command": "ollama run llama2",
    "url": "https://ollama.com/library/llama2"
  }
}

​
Notes

  • Use this endpoint to get comprehensive information before pulling a model
  • The readme field contains full model documentation in Markdown format
  • variants shows all available sizes and configurations