curl http://localhost:11434/api/create -d '{ "from": "gemma4", "model": "alpaca", "system": "You are Alpaca, a helpful AI assistant. You only answer with Emojis."}'
{ "status": "success"}
Body
application/json
model
string
required
Name for the model to create
from
string
Existing model to create from
template
string
Prompt template to use for the model
renderer
string
Name of the renderer for the model
parser
string
Name of the parser for the model
files
object
Source file names mapped to their SHA-256 digests. Split GGUF models must include each shard under its original split filename.
Show child attributes
draft_files
object
Draft source file names mapped to their SHA-256 digests
Show child attributes
license
stringstring[]
License string or list of licenses for the model
system
string
System prompt to embed in the model
parameters
object
Key-value parameters for the model
messages
object[]
Message history to use for the model
Show child attributes
quantize
string
Quantization level to apply during import (e.g. nvfp4)
draft_quantize
string
Quantization level to apply to draft weights during import
requires
string
Minimum Ollama version required by the model
capabilities
string[]
Capabilities to add without removing inherited or inferred capabilities (e.g. decision for compatible decision models)
stream
boolean
default:true
Stream status updates
Response
Stream of create status updates
status
string
Current status message
⌘I
curl http://localhost:11434/api/create -d '{ "from": "gemma4", "model": "alpaca", "system": "You are Alpaca, a helpful AI assistant. You only answer with Emojis."}'