> For the complete documentation index, see [llms.txt](https://docs.otari.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.otari.ai/self-hosting/providers/openai.md). # Connect OpenAI Serve OpenAI models (GPT-4o, GPT-4.1, the o-series, and others) through Otari with your OpenAI API key. See any-llm's [supported providers](https://docs.mozilla.ai/any-llm/providers/) for the current provider list and [Models](https://docs.otari.ai/reference/gateway/models) for Otari's model-name format. ## Before you start * A running Otari with `OTARI_SECRET_KEY` set, and its master key. The [Self-hosted quickstart](/self-hosting/self-hosted-quickstart.md) sets up both. * An OpenAI API key from . ## Add the key Add OpenAI as a hosted provider, on **Hosted providers** in the dashboard's deployment view or through the API: ```bash curl -X POST http://localhost:8000/api/v1/hosted-providers \ -H "Authorization: Bearer $OTARI_MASTER_KEY" \ -H "Content-Type: application/json" \ -d '{"provider": "openai", "api_key": "sk-..."}' ``` Otari stores the key encrypted and offers every model OpenAI lists on it; each model with a price is switched on, and an unpriced model arrives switched off. To keep the provider in configuration instead, declare it under `hosted_providers:` in `config.yml` (see [Hosted providers in config.yml](/self-hosting/configuration.md#hosted-providers-in-configyml)). Otari does not read `OPENAI_API_KEY` from its environment unless a declaration names it; see [Configuring a provider](https://docs.otari.ai/reference/gateway/models#configuring-a-provider). ### Optional settings The same body takes `api_base`, for a proxy or another OpenAI-compatible endpoint, and `client_args`, passed through to the underlying `any-llm` provider client for options such as custom headers or timeouts: ```json { "provider": "openai", "api_key": "sk-...", "client_args": { "timeout": 60, "custom_headers": {"OpenAI-Organization": "org-..."} } } ``` ## Choose a model Models are addressed as `openai:`: ``` openai:gpt-4o openai:gpt-4o-mini openai:o4-mini ``` Everything after the colon is passed to OpenAI. A model is served once it is switched on under the provider on the Hosted providers page. ## Verify Make a request with your Otari API key (`tk-...`) through the OpenAI client (`pip install openai`): ```python from openai import OpenAI client = OpenAI(base_url="http://localhost:8000/api/v1", api_key="tk-...") resp = client.chat.completions.create( model="openai:gpt-4o-mini", messages=[{"role": "user", "content": "Say hello in five words."}], ) print(resp.choices[0].message.content) ``` Expected (sample) output: ``` Hello, nice to meet you! ``` ## Troubleshooting | Symptom | Likely cause | | ------------------------------------------------------------------------------------------------------------------------------------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | HTTP 401 `Invalid master key` when adding the provider | You sent an API key (`tk-...`) or the wrong master key. `/api/v1/hosted-providers` takes the master key. | | HTTP 424 (code `provider_not_configured`) | Nothing serves the model: no OpenAI provider is configured, it is switched off, or the model is not switched on under it. Switch the model on under OpenAI on the Hosted providers page. | | HTTP 424 `The provider rejected the gateway's credentials` (code `provider_credential_rejected`) | OpenAI rejected the key Otari sent as invalid or lacking access. Replace it where the provider is configured: on the Hosted providers page, or in its `hosted_providers:` entry (or the variable that entry names, such as `OPENAI_API_KEY`) in `config.yml`. | | HTTP 404 with OpenAI's own "model not found" text | The name after `openai:` is misspelled or not a model your key can access; the detail carries OpenAI's message. `The requested model was not found on the provider` appears instead when OpenAI sent no message. | | HTTP 502 `The upstream provider account is out of credit or over its billing limit.` | The OpenAI account has no credit left or hit its billing hard limit. Top it up, or point the model at a provider that has credit. The usage log keeps the status OpenAI returned. | | HTTP 502 `The provider failed the request with HTTP 5xx`, `The provider closed the connection before answering`, or `The gateway could not connect to the provider` | OpenAI failed on its side or the connection broke. These are retryable. `LLM provider error` is the same answer for a failure Otari could not place. | | HTTP 402 `No pricing configured for model ...` | The model has no price and `require_pricing` rejected the request. Set a price for it (below), or enable `default_pricing: true`. | ## Pricing (optional) Set a model's price on the Hosted providers page, or add a `pricing:` entry in `config.yml`: ```yaml pricing: openai:gpt-4o-mini: input_price_per_million: 0.15 output_price_per_million: 0.60 ``` See [Configuration](/self-hosting/configuration.md) for all options. --- # Agent Instructions This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com. ## Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation by asking a question. Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters: ``` GET https://docs.otari.ai/self-hosting/providers/openai.md?ask=&goal= ``` `ask` is the immediate question: it should be specific, self-contained, and written in natural language. `goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.