Documentation
Connect to the API, configure an AI client, or create a video.
Connect to the API, configure an AI client, or create a video.
OpenAI-compatible API
https://api.unlimgpt.com/v1Anthropic-compatible API
https://api.unlimgpt.comCreate a key and send your first Responses request.
Configure a supported client using the CLI or a guided setup.
Open the video model catalog and prepare text or reference media.
Set BASE_URL to your HTTPS API origin without a trailing slash or /v1, and pass API_KEY from your credential store. OpenAI-compatible SDKs use ${BASE_URL}/v1; Anthropic-compatible SDKs use BASE_URL.
Chat and API use one account; API access depends on the plan and key permissions. Configure your client with its protocol's base URL, not a single request URL:
| Client or protocol | Base URL | Path the client calls |
|---|---|---|
| OpenAI Responses and Chat Completions | ${BASE_URL}/v1 | /responses or /chat/completions |
| Anthropic Messages | BASE_URL | /v1/messages |
Pass the raw key only through the client's supported credential setting; never append it to the URL. The compatibility matrix shows the exact setup mode for each AI client.
After creating an API key, set API_KEY in your shell from a secure store. Choose a model ID returned by the free model catalog; replace the sample ID if your account lists a different model. This request can consume credits.
printf 'Authorization: Bearer %s\n' "$API_KEY" \
| curl -q -H @- "$BASE_URL/v1/responses" \
-H "Content-Type: application/json" \
-d '{
"model": "gpt-5.6-sol",
"input": "Say hello in one sentence.",
"max_output_tokens": 64,
"stream": false,
"store": false
}'
The response contains id, status, model, output_text, structured output, and usage. The text is also available in an output[].content[] block of type output_text. This example explicitly avoids storing a retrievable response; activity metadata is still recorded. Check the recent operation's endpoint, time and outcome in API Activity. Its row ID differs from the HTTP request-id header. See the complete first-request guide before sending it.