Assistants and agents
Connect supported chat models through an OpenAI-compatible endpoint. Use a session ID when successive turns benefit from route affinity.
Start with a scene in Creator: choose a video model, describe your idea, and review the generated clip. Download the result for your project.
Use one platform for access and billing while choosing the endpoint and model that fit each part of your product.
Connect supported chat models through an OpenAI-compatible endpoint. Use a session ID when successive turns benefit from route affinity.
Combine text, image, video, and audio in one workflow. Check each model’s inputs and track asynchronous media tasks through their status endpoints.
Give teams access to approved models and use key ownership and usage records to see which services are making requests.
Token360 resolves the public model ID to an eligible upstream route. Model compatibility, route health, account policy, and available capacity affect that decision.
Use the model name from the catalog in your request. Confirm the selected endpoint and accepted fields before switching models.
Use a stable session ID to prefer an eligible upstream route across turns. Health and policy checks still apply.
Check the model schema for fallback controls and follow the retry guidance for timeouts and rate limits.
Set it as TOKEN360_API_KEY in your local environment.
This example uses claude-opus-5. You can swap in another catalog model that supports Chat Completions.
Check the response and usage. For image, video, or audio, use the endpoint listed on the model page.
python3 -c 'import json; print(json.dumps({
"model": "claude-opus-5",
"messages": [{"role": "user", "content": "Reply with hello."}]
}))' > gateway-request.json &&
curl --fail-with-body https://api.token360.ai/v1/chat/completions \
-H "Authorization: Bearer ${TOKEN360_API_KEY:?Set your API key}" \
-H "Content-Type: application/json" \
--data-binary @gateway-request.jsonModel usage is billed according to its Token360 pricing.
Use enterprise sub-accounts, unified wallet billing, and tenant-wide API key audit as more teams put AI into production.
Review key metadata across your tenant and revoke keys when access changes.
Set model access policies for members and confirm which routes meet your workload’s requirements.
Manage team usage under a unified wallet. Discuss volume, concurrency, and commercial terms with the enterprise team.
An LLM gateway focuses on language model access. A multimodal AI gateway also handles workloads such as image, video, and audio generation. Read the AI gateway guide.
For supported Chat Completions workloads, set the base URL to https://api.token360.ai/v1 and use a Token360 key and supported model ID. Check the model page for provider-specific fields.
Eligible routes depend on model configuration, account policy, and request compatibility. Check the model’s published schema and the routing guide for available fallback controls.
Rates and billing units depend on the model and request settings. See the Token360 model catalog for the applicable price and meter.
Choose a model, run a representative request, and check the result and usage before integrating it into your product.