Load an adapter

curl --request POST \
  --url https://api.fireworks.ai/v1/accounts/{account_id}/deployedModels \
  --header 'Content-Type: application/json' \
  --data '
{
  "model": "<string>",
  "serverless": true,
  "deployment": "<string>",
  "displayName": "<string>",
  "description": "<string>"
}
'

import requests

url = "https://api.fireworks.ai/v1/accounts/{account_id}/deployedModels"

payload = {
    "model": "<string>",
    "serverless": True,
    "deployment": "<string>",
    "displayName": "<string>",
    "description": "<string>"
}
headers = {"Content-Type": "application/json"}

response = requests.post(url, json=payload, headers=headers)

print(response.text)

const options = {
  method: 'POST',
  headers: {'Content-Type': 'application/json'},
  body: JSON.stringify({
    model: '<string>',
    serverless: true,
    deployment: '<string>',
    displayName: '<string>',
    description: '<string>'
  })
};

fetch('https://api.fireworks.ai/v1/accounts/{account_id}/deployedModels', options)
  .then(res => res.json())
  .then(res => console.log(res))
  .catch(err => console.error(err));

<?php

$curl = curl_init();

curl_setopt_array($curl, [
  CURLOPT_URL => "https://api.fireworks.ai/v1/accounts/{account_id}/deployedModels",
  CURLOPT_RETURNTRANSFER => true,
  CURLOPT_ENCODING => "",
  CURLOPT_MAXREDIRS => 10,
  CURLOPT_TIMEOUT => 30,
  CURLOPT_HTTP_VERSION => CURL_HTTP_VERSION_1_1,
  CURLOPT_CUSTOMREQUEST => "POST",
  CURLOPT_POSTFIELDS => json_encode([
    'model' => '<string>',
    'serverless' => true,
    'deployment' => '<string>',
    'displayName' => '<string>',
    'description' => '<string>'
  ]),
  CURLOPT_HTTPHEADER => [
    "Content-Type: application/json"
  ],
]);

$response = curl_exec($curl);
$err = curl_error($curl);

curl_close($curl);

if ($err) {
  echo "cURL Error #:" . $err;
} else {
  echo $response;
}

package main

import (
	"fmt"
	"strings"
	"net/http"
	"io"
)

func main() {

	url := "https://api.fireworks.ai/v1/accounts/{account_id}/deployedModels"

	payload := strings.NewReader("{\n  \"model\": \"<string>\",\n  \"serverless\": true,\n  \"deployment\": \"<string>\",\n  \"displayName\": \"<string>\",\n  \"description\": \"<string>\"\n}")

	req, _ := http.NewRequest("POST", url, payload)

	req.Header.Add("Content-Type", "application/json")

	res, _ := http.DefaultClient.Do(req)

	defer res.Body.Close()
	body, _ := io.ReadAll(res.Body)

	fmt.Println(string(body))

}

HttpResponse<String> response = Unirest.post("https://api.fireworks.ai/v1/accounts/{account_id}/deployedModels")
  .header("Content-Type", "application/json")
  .body("{\n  \"model\": \"<string>\",\n  \"serverless\": true,\n  \"deployment\": \"<string>\",\n  \"displayName\": \"<string>\",\n  \"description\": \"<string>\"\n}")
  .asString();

require 'uri'
require 'net/http'

url = URI("https://api.fireworks.ai/v1/accounts/{account_id}/deployedModels")

http = Net::HTTP.new(url.host, url.port)
http.use_ssl = true

request = Net::HTTP::Post.new(url)
request["Content-Type"] = 'application/json'
request.body = "{\n  \"model\": \"<string>\",\n  \"serverless\": true,\n  \"deployment\": \"<string>\",\n  \"displayName\": \"<string>\",\n  \"description\": \"<string>\"\n}"

response = http.request(request)
puts response.read_body

POST

accounts

{account_id}

deployedModels

Load an adapter

curl --request POST \
  --url https://api.fireworks.ai/v1/accounts/{account_id}/deployedModels \
  --header 'Content-Type: application/json' \
  --data '
{
  "model": "<string>",
  "serverless": true,
  "deployment": "<string>",
  "displayName": "<string>",
  "description": "<string>"
}
'

import requests

url = "https://api.fireworks.ai/v1/accounts/{account_id}/deployedModels"

payload = {
    "model": "<string>",
    "serverless": True,
    "deployment": "<string>",
    "displayName": "<string>",
    "description": "<string>"
}
headers = {"Content-Type": "application/json"}

response = requests.post(url, json=payload, headers=headers)

print(response.text)

const options = {
  method: 'POST',
  headers: {'Content-Type': 'application/json'},
  body: JSON.stringify({
    model: '<string>',
    serverless: true,
    deployment: '<string>',
    displayName: '<string>',
    description: '<string>'
  })
};

fetch('https://api.fireworks.ai/v1/accounts/{account_id}/deployedModels', options)
  .then(res => res.json())
  .then(res => console.log(res))
  .catch(err => console.error(err));

<?php

$curl = curl_init();

curl_setopt_array($curl, [
  CURLOPT_URL => "https://api.fireworks.ai/v1/accounts/{account_id}/deployedModels",
  CURLOPT_RETURNTRANSFER => true,
  CURLOPT_ENCODING => "",
  CURLOPT_MAXREDIRS => 10,
  CURLOPT_TIMEOUT => 30,
  CURLOPT_HTTP_VERSION => CURL_HTTP_VERSION_1_1,
  CURLOPT_CUSTOMREQUEST => "POST",
  CURLOPT_POSTFIELDS => json_encode([
    'model' => '<string>',
    'serverless' => true,
    'deployment' => '<string>',
    'displayName' => '<string>',
    'description' => '<string>'
  ]),
  CURLOPT_HTTPHEADER => [
    "Content-Type: application/json"
  ],
]);

$response = curl_exec($curl);
$err = curl_error($curl);

curl_close($curl);

if ($err) {
  echo "cURL Error #:" . $err;
} else {
  echo $response;
}

package main

import (
	"fmt"
	"strings"
	"net/http"
	"io"
)

func main() {

	url := "https://api.fireworks.ai/v1/accounts/{account_id}/deployedModels"

	payload := strings.NewReader("{\n  \"model\": \"<string>\",\n  \"serverless\": true,\n  \"deployment\": \"<string>\",\n  \"displayName\": \"<string>\",\n  \"description\": \"<string>\"\n}")

	req, _ := http.NewRequest("POST", url, payload)

	req.Header.Add("Content-Type", "application/json")

	res, _ := http.DefaultClient.Do(req)

	defer res.Body.Close()
	body, _ := io.ReadAll(res.Body)

	fmt.Println(string(body))

}

HttpResponse<String> response = Unirest.post("https://api.fireworks.ai/v1/accounts/{account_id}/deployedModels")
  .header("Content-Type", "application/json")
  .body("{\n  \"model\": \"<string>\",\n  \"serverless\": true,\n  \"deployment\": \"<string>\",\n  \"displayName\": \"<string>\",\n  \"description\": \"<string>\"\n}")
  .asString();

require 'uri'
require 'net/http'

url = URI("https://api.fireworks.ai/v1/accounts/{account_id}/deployedModels")

http = Net::HTTP.new(url.host, url.port)
http.use_ssl = true

request = Net::HTTP::Post.new(url)
request["Content-Type"] = 'application/json'
request.body = "{\n  \"model\": \"<string>\",\n  \"serverless\": true,\n  \"deployment\": \"<string>\",\n  \"displayName\": \"<string>\",\n  \"description\": \"<string>\"\n}"

response = http.request(request)
puts response.read_body

Preview Loads a LoRA adapter onto the shared serverless pool. Loading takes about 2 to 5 minutes. You’re billed per token only when you call the adapter.

curl -X POST "https://api.fireworks.ai/v1/accounts/$ACCOUNT_ID/deployedModels" \
  -H "Authorization: Bearer $FIREWORKS_API_KEY" \
  -H "Content-Type: application/json" \
  -d "{\"model\": \"accounts/$ACCOUNT_ID/models/my-lora\", \"serverless\": true}"

The firectl equivalent:

firectl model load-lora my-lora --serverless --wait

Path parameters

string

required

Your account ID.

Body

string

required

The LoRA model to load: accounts/{account_id}/models/{adapter_id}. It must be in the READY state, trained on a supported base model, with rank 64 or lower.

boolean

true to load onto Fireworks’ shared serverless pool. Required unless you set deployment.

string

Load onto one of your dedicated deployments instead: accounts/{account_id}/deployments/{deployment_id}. The deployment must have addons enabled. Billed by GPU-hour, not per token.

string

Optional display name.

string

Optional description.

Set exactly one of serverless: true or deployment. A request with both, or neither, is rejected.

Response

Returns the deployed model. It starts in DEPLOYING; send traffic once it’s DEPLOYED. Save name: its last segment is the deployed model ID you need to unload over REST.

{
  "name": "accounts/your-account-id/deployedModels/abc123",
  "model": "accounts/your-account-id/models/my-lora",
  "serverless": true,
  "default": true,
  "state": "DEPLOYING"
}

Behavior

One serverless copy per adapter. Loading the same adapter onto serverless again fails while the first copy exists.
The first request is slower. The first request after loading a LoRA adapter can take up to a minute. Later requests have much lower latency.
The bare model name. The first copy of an adapter you load becomes its default and answers accounts/<ACCOUNT_ID>/models/<LORA_ID>. If another copy already holds that name, this copy loads with default: false. Call it with accounts/<ACCOUNT_ID>/deployedModels/<DEPLOYED_MODEL_ID>. There is no #serverless suffix. See Which model string to use.
Adapter limit. You can load up to 3 adapters per base model on serverless at a time. Going over it returns 429 RESOURCE_EXHAUSTED. Unload an adapter to free up a slot. For more slots, contact inquiries@fireworks.ai or your account team. See Supported models and limits.

Errors

Status	Cause	Fix
`400`	Rank above 64	Retrain with `--lora-rank 64` or lower
`400`	Base model isn’t supported on serverless	Train on `accounts/fireworks/models/glm-5p3-flash`
`400`	`serverless` and `deployment` both set or both missing	Send only `"serverless": true`
`404`	The adapter doesn’t exist in this account	Check the adapter ID with `firectl model get my-lora`
`409`	Already loaded, or an unload is still in progress	Check `firectl deployed-model list` and wait for the unload to finish
`429`	Adapter limit reached	Unload an adapter you don’t need

Load onto a dedicated deployment instead

curl -X POST "https://api.fireworks.ai/v1/accounts/$ACCOUNT_ID/deployedModels" \
  -H "Authorization: Bearer $FIREWORKS_API_KEY" \
  -H "Content-Type: application/json" \
  -d "{\"model\": \"accounts/$ACCOUNT_ID/models/my-lora\", \"deployment\": \"accounts/$ACCOUNT_ID/deployments/$DEPLOYMENT_ID\"}"

⌘I

API Reference

Inference

Deployments

Training

Training SDK

Evals

Admin

Load an adapter

Path parameters

Body

Response

Behavior

Errors

Load onto a dedicated deployment instead

​Path parameters

​Body

​Response

​Behavior

​Errors

​Load onto a dedicated deployment instead

Path parameters

Body

Response

Behavior

Errors

Load onto a dedicated deployment instead