Skip to main content
Generate a response for a given prompt with a provided model. This is the core text generation endpoint that supports completions, image generation, and multimodal inputs.

​
Request

​
Endpoint

POST /api/generate

​
Request Body

​
model
string
required
Model name to use for generation (e.g., llama3.2, mistral, stable-diffusion)
​
prompt
string
required
The prompt to generate a response for
​
suffix
string
Text that comes after the inserted text (for fill-in-the-middle models)
​
system
string
System message to override the model’s default system prompt
​
template
string
Custom prompt template to override the model’s default
​
context
array
Context from a previous generation to maintain conversational memory (deprecated)
​
stream
boolean
default:"true"
Enable streaming of response chunks. Set to false to wait for complete response.
​
raw
boolean
default:"false"
If true, no formatting will be applied to the prompt
​
format
string | object
Format to return response in. Use "json" for JSON mode or provide a JSON schema.
​
images
array
Array of base64-encoded images for multimodal models
​
options
object
Model-specific options for customizing generation
​
keep_alive
string | number
default:"5m"
Duration to keep model in memory (e.g., "5m", "1h", -1 for indefinite)
​
think
boolean | string
Enable thinking mode for reasoning models (true/false or "high", "medium", "low")
​
truncate
boolean
default:"true"
Truncate prompt if it exceeds context length
​
shift
boolean
default:"true"
Shift context when hitting context length instead of erroring
​
logprobs
boolean
default:"false"
Return log probabilities for output tokens
​
top_logprobs
integer
default:"0"
Number of top tokens to return at each position (0-20, requires logprobs: true)

​
Image Generation Parameters

The following parameters are only used with image generation models:
​
width
integer
Width of generated image in pixels (max 4096)
​
height
integer
Height of generated image in pixels (max 4096)
​
steps
integer
Number of diffusion steps for image generation

​
Response

​
Response Fields

​
model
string
The model used for generation
​
created_at
string
Timestamp in ISO 8601 format
​
response
string
The generated text response
​
thinking
string
Reasoning text when thinking mode is enabled
​
done
boolean
Whether generation is complete
​
done_reason
string
Reason for completion: stop, length, load, unload
​
context
array
Token encoding of the conversation for maintaining memory (deprecated)
​
total_duration
integer
Total duration in nanoseconds
​
load_duration
integer
Model loading time in nanoseconds
​
prompt_eval_count
integer
Number of tokens in the prompt
​
prompt_eval_duration
integer
Time evaluating the prompt in nanoseconds
​
eval_count
integer
Number of tokens generated
​
eval_duration
integer
Time generating response in nanoseconds
​
logprobs
array
Log probability information (when enabled)
​
image
string
Base64-encoded generated image (image generation models only)
​
completed
integer
Number of completed steps during image generation
​
total
integer
Total steps for image generation

​
Examples

​
Basic Text Generation

curl http://localhost:11434/api/generate -d '{
  "model": "llama3.2",
  "prompt": "What is the capital of France?",
  "stream": false
}'

​
Example Response

{
  "model": "llama3.2",
  "created_at": "2024-02-24T12:34:56.789Z",
  "response": "The capital of France is Paris.",
  "done": true,
  "done_reason": "stop",
  "context": [128000, 3923, 374, 279, ...],
  "total_duration": 2345678901,
  "load_duration": 123456789,
  "prompt_eval_count": 8,
  "prompt_eval_duration": 234567890,
  "eval_count": 8,
  "eval_duration": 1987654321
}

​
Streaming Generation

import requests
import json

response = requests.post(
    'http://localhost:11434/api/generate',
    json={'model': 'llama3.2', 'prompt': 'Write a haiku'},
    stream=True
)

for line in response.iter_lines():
    if line:
        chunk = json.loads(line)
        print(chunk['response'], end='', flush=True)

​
JSON Mode

curl http://localhost:11434/api/generate -d '{
  "model": "llama3.2",
  "prompt": "List 3 colors",
  "format": "json",
  "stream": false
}'

​
Multimodal Input (Vision Models)

curl http://localhost:11434/api/generate -d '{
  "model": "llama3.2-vision",
  "prompt": "What is in this image?",
  "images": ["<base64-encoded-image>"],
  "stream": false
}'

​
Error Responses

​
error
string
Description of the error

​
Common Errors

  • 400 Bad Request: Invalid model name, parameters, or image dimensions
  • 404 Not Found: Model not found
  • 500 Internal Server Error: Generation error
To unload a model from memory, send an empty prompt with keep_alive: 0.

Build docs developers (and LLMs) love

Get started for freeTalk to us