One OpenAI base URL

Use https://api.tekiz.ai/v1 for hosted routing, http://127.0.0.1:8790/v1 for local routing, or a Tailnet host when provider credentials should stay private.

Use more than one OpenAI key

Add multiple OpenAI API keys. TEKIZ.AI spreads requests across keys that are working and pauses only the key that hits a rate limit or retryable error.

Works with Chat Completions and Responses

Use OpenAI-compatible Chat Completions or Responses with stable names such as tekizai/frontier. You can change the models behind that name. Every client keeps the same model alias.

Set up OpenAI router setup

export OPENAI_BASE_URL=https://api.tekiz.ai/v1
export OPENAI_API_KEY=sk_tekiz_your_generated_key

curl "$OPENAI_BASE_URL/chat/completions" \
  -H "Authorization: Bearer $OPENAI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "tekizai/frontier",
    "messages": [{"role": "user", "content": "Route this through the best healthy model"}]
  }'

Keep working when an account is busy

After a retryable rate limit, TEKIZ.AI can use another key, provider, or local model. Your app does not need its own chain of retries.

Send files to services that accept them

Images, audio, video, and documents are sent only to models that can handle that kind of input.

Use accounts you are allowed to use

TEKIZ.AI provides routing, API keys, usage limits, and account history. It does not bypass provider terms or grant access to accounts you do not own.

Why put TEKIZ.AI in front of the OpenAI SDK?

Your tools keep the interface they already use. TEKIZ.AI handles model choice, checks which keys are working, sends files to compatible models, and tries an available backup without requiring changes in every app.

Next steps

Create an API key, copy the hosted quickstart, read the API reference, or compare OpenRouter.

Stay within your approved services

If your company approves only OpenAI, keep that boundary. TEKIZ.AI can reduce token burn while making better use of existing subscriptions and account billing.