Documentation

Integrate production inference in minutes. Choose a compatible API, create a key, and send your first request.

Quickstart

Every request to the inference API uses a token created in the console. Tokens are sent in the Authorization header using the Bearer scheme. Model IDs come from Dashboard → Models.

API origin for this deployment

Base URL

https://inferenceapi.wethinkdigital.solutions/v1

Resolved from NEXT_PUBLIC_API_BASE_URL and the /v1 path. Configure this in the environment before deploying.

Send a non-streaming chat completion request:

quickstart

curl https://inferenceapi.wethinkdigital.solutions/v1/chat/completions \
  -H "Authorization: Bearer sk_live_xxxxxxxx" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "model-pro",
    "messages": [
      { "role": "user", "content": "Explain idempotency in one sentence." }
    ]
  }'

Create your token first

The placeholder sk_live_xxxxxxxx will not authenticate. Generate a real token from Dashboard → Tokens. The secret is shown once.

Base URL

All inference endpoints live under a versioned base URL. Use the origin provided by your deployment, followed by /v1.

base-url

https://inferenceapi.wethinkdigital.solutions/v1
https://inferenceapi.wethinkdigital.solutions/v1/chat/completions
https://inferenceapi.wethinkdigital.solutions/v1/messages
https://inferenceapi.wethinkdigital.solutions/v1/models
https://inferenceapi.wethinkdigital.solutions/v1/usage
https://inferenceapi.wethinkdigital.solutions/v1/balance

The console itself talks to a separate, internal management API. You should not use console URLs from your application. Only the inference base URL and API tokens are relevant to product traffic.

API compatibility

Inference Cloud exposes two compatible surfaces over the same catalog and billing balance. Use whichever matches your existing stack.

OpenAI-compatible

Chat completions, streaming, tool calls, and model listing under /v1/chat/completions and /v1/models.

OpenAI guide

Anthropic-compatible

Messages with content blocks, streaming, and model metadata under /v1/messages.

Anthropic guide

Guides

AuthenticationCreate, use, rotate, and revoke API tokens. Understand the difference between console sessions and API keys.OpenAI-compatible APIChat completions with streaming, model listing, and examples using the OpenAI SDK with a base URL override.Anthropic-compatible APIMessages API with x-api-key and anthropic-version headers, streaming, and SDK examples.Usage & balanceQuery usage, check your balance, understand per-token cost semantics, errors, and rate limits.SecurityKeep API keys server-side, rotate credentials safely, and design integrations that do not leak secrets.

Errors

Non-2xx responses use a consistent JSON envelope. Always branch on the HTTP status first, then use error.code for programmatic handling.

POSThttps://inferenceapi.wethinkdigital.solutions/v1/chat/completions

error-response.json

{
  "error": {
    "code": "insufficient_credit",
    "message": "Your credit balance is too low for this request.",
    "requestId": "req_01J..."
  }
}

See Usage & balance for the complete error and rate-limit reference.

All documentation Create an API token →