Crusoe

Crusoe Managed Inference exposes an OpenAI-compatible API for open-weight models such as meta-llama/Llama-3.3-70B-Instruct, deepseek-ai/DeepSeek-V3-0324, and openai/gpt-oss-120b. Use Custom Endpoints in the AI Gateway to register Crusoe once, keep the Crusoe API key on the AI Gateway, and let your applications call models through a single TrueFoundry API key with access control and tracing.

Prerequisites

Create a Crusoe Cloud account

Generate a Crusoe API key

In the Crusoe Cloud Console, go to Admin → Security → Intelligence API keys and click Create.Store the key securely. You will paste it into the AI Gateway’s Custom Headers configuration — not in client application code.

Get your TrueFoundry API key and gateway URL

You need a TrueFoundry API key (TFY_API_KEY) and your gateway base URL to call models through the AI Gateway. See Gateway base URL and Authentication.

Adding models

Add Crusoe to the AI Gateway using Custom Endpoints.

Create a Custom Endpoint model account

In the TrueFoundry dashboard, go to AI Gateway → Models → Custom Endpoints and click Add Custom Endpoint.In Configure Account, set:

Name (Account Name): crusoe — this is your model account name and appears as the first path segment in every request URL ({providerAccountName})
Endpoint Type: None
Header Auth: keep disabled

Add collaborators so other users or teams can use this account. See Gateway access control.

Add a Crusoe endpoint

On the Endpoints step, add an integration and configure:

Display Name: crusoe_managed_inference — this is your custom endpoint display name and appears as the second path segment in every request URL ({endpointName})
Base URL: https://api.inference.crusoecloud.com (no trailing slash)

Enable Custom Headers and add:

Authorization: Bearer <CRUSOE_API_KEY>
Content-Type: application/json

Keep Header Auth and TLS Settings disabled unless your deployment requires them.

The Authorization header here is sent from the AI Gateway to Crusoe. Your application should only send the TrueFoundry API key to the AI Gateway.

Set access control and save

On the Access Control step, choose who can manage and use this model account, then Save.

Manager: users/teams who can edit or delete the custom endpoint
User: users/teams who can call the endpoint (for example, everyone)

Inference

Once saved, call Crusoe through the AI Gateway’s proxy-api path. URL shape and path rules are documented under Custom Endpoints.

How the request URL is built

The AI Gateway URL has two values you configure in the dashboard — they are not Crusoe model IDs:

URL segment	Dashboard field	Example in this guide
`{providerAccountName}`	Account Name (Configure Account → Name)	`crusoe`
`{endpointName}`	Custom endpoint Display Name (Endpoints step)	`crusoe_managed_inference`

Full URL pattern:

{GATEWAY_BASE_URL}/proxy-api/{providerAccountName}/{endpointName}/v1/chat/completions

With the example configuration above:

{GATEWAY_BASE_URL}/proxy-api/crusoe/crusoe_managed_inference/v1/chat/completions

Segment	Example value	Meaning
`{GATEWAY_BASE_URL}`	`https://internal.devtest.truefoundry.tech/api/llm`	Your AI Gateway base URL
`crusoe`	Account Name you set in Configure Account	`{providerAccountName}`
`crusoe_managed_inference`	Display Name you set for the custom endpoint	`{endpointName}`
`v1/chat/completions`	Upstream path	Appended to the integration Base URL

The AI Gateway forwards the request to:

https://api.inference.crusoecloud.com/v1/chat/completions

Choosing a Crusoe model

Crusoe hosts open-weight models such as meta-llama/Llama-3.3-70B-Instruct, deepseek-ai/DeepSeek-V3-0324, and openai/gpt-oss-120b. You do not need a separate custom endpoint per model — set the model field in the JSON request body to the Crusoe model ID you want. See the Crusoe serverless inference docs for the full list.

Replace crusoe and crusoe_managed_inference in the URL with your own Account Name and Display Name if you used different values during setup.

Supported APIs

API	Endpoint	Tracing	Cost Tracking
Chat Completions	`/proxy-api/{providerAccountName}/{endpointName}/v1/chat/completions`	✅

Chat Completions

Before you start: Replace {GATEWAY_BASE_URL} with your gateway base URL (how to find it) and set TFY_API_KEY to your TrueFoundry API key.

Request headers (client → gateway)

Header	Value
`Authorization`	`Bearer <TFY_API_KEY>`
`Content-Type`	`application/json`
`Accept-Encoding`	`identity`

OpenAI Python SDK

Because Crusoe is OpenAI-compatible, you can point the OpenAI SDK at the AI Gateway proxy path:

import os

from openai import OpenAI

GATEWAY_BASE_URL = "{GATEWAY_BASE_URL}"  # e.g. https://internal.devtest.truefoundry.tech/api/llm
TFY_API_KEY = os.environ.get("TFY_API_KEY", "your-tfy-api-key")

PROVIDER_ACCOUNT_NAME = "crusoe"  # Account Name
ENDPOINT_NAME = "crusoe_managed_inference"  # Custom endpoint Display Name

client = OpenAI(
    api_key=TFY_API_KEY,
    base_url=f"{GATEWAY_BASE_URL}/proxy-api/{PROVIDER_ACCOUNT_NAME}/{ENDPOINT_NAME}/v1",
)

response = client.chat.completions.create(
    model="meta-llama/Llama-3.3-70B-Instruct",
    messages=[
        {"role": "system", "content": "You are a helpful assistant."},
        {"role": "user", "content": "Tell me about TrueFoundry AI Gateway"},
    ],
)

print(response.choices[0].message.content)

cURL

curl -X POST "{GATEWAY_BASE_URL}/proxy-api/crusoe/crusoe_managed_inference/v1/chat/completions" \
  -H "Authorization: Bearer $TFY_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Accept-Encoding: identity" \
  -d '{
    "model": "meta-llama/Llama-3.3-70B-Instruct",
    "messages": [
      {"role": "user", "content": "Tell me about TrueFoundry AI Gateway"}
    ]
  }'

Support scope: Custom Endpoints proxy requests transparently to Crusoe. See the Crusoe Managed Inference docs for supported models and request fields. For gateway limitations on Custom Endpoints (HTTPS, streaming, and so on), see Custom Endpoints.

Get Started

LLM Gateway

MCP Registry and Gateway

Skills Registry

Prompt Registry

Guardrails and Security

Observability

Deployment

Admin Guide

Chat

Messages

Embeddings

Rerank

Responses

Image

Audio

Batch

Files

Fine-tuning

Moderations

Video

Models

Prerequisites

Adding models

Inference

How the request URL is built

Choosing a Crusoe model

Supported APIs

Request headers (client → gateway)

OpenAI Python SDK

cURL

​Prerequisites

​Adding models

​Inference

​How the request URL is built

​Choosing a Crusoe model

​Supported APIs

​Request headers (client → gateway)

​OpenAI Python SDK

​cURL

Prerequisites

Adding models

Inference

How the request URL is built

Choosing a Crusoe model

Supported APIs

Request headers (client → gateway)

OpenAI Python SDK

cURL