OpenAI

How to configure OpenAI in Aimogen Pro, which models and features it unlocks, organisation IDs, the Responses API, and the OpenAI-only capabilities.

OpenAI is the provider with the broadest feature coverage in Aimogen Pro. Several plugin features work only with an OpenAI key.

Get a key#

Sign in at platform.openai.com.
Open API keys and create a secret key.
Add a payment method under Billing. New accounts without credit return insufficient_quota on the first request.

Configure#

Aimogen Pro › Settings › API Keys

Setting	Value
Main API Service Provider Selector	`openai`
OpenAI / AiomaticAPI API Keys	One key per line
OpenAI Organization ID (Optional)	Only if your account belongs to several organisations

Select Test API key, then save.

Organization ID. If your OpenAI account belongs to multiple organisations, this header decides which one is billed. Leave it empty otherwise — an incorrect value produces a permission error on every request.

Aimogen Pro > Settings > API Keys

Models#

Aimogen Pro 2.8.8 ships with 60 OpenAI chat models in its list, including the GPT-5.x family, the GPT-4.x family, the reasoning o-series and a Codex model.

Family	Examples	Typical use
GPT-5.6	`gpt-5.6-luna`, `gpt-5.6-terra`, `gpt-5.6-sol`	Current flagship tier
GPT-5.5 / 5.4 / 5.2 / 5.1 / 5	`gpt-5.5`, `gpt-5.5-pro`, `gpt-5.4-mini`, `gpt-5-mini`, `gpt-5-nano`	General writing
GPT-4.1	`gpt-4.1`, `gpt-4.1-mini`, `gpt-4.1-nano`	Cost-sensitive bulk work
GPT-4o	`gpt-4o`, `gpt-4o-mini`, plus search-preview variants	Vision and general use
Reasoning	`o1`, `o1-pro`, `o3`, `o3-pro`, `o3-mini`, `o4-mini`	Multi-step analysis
Coding	`gpt-5.3-codex`	Code generation, Snippets

The plugin default is gpt-5-mini (AIMOGEN_DEFAULT_MODEL).

Deprecated models are removed on each release. If a rule stops working right after an update with model_not_found, that is why — see Updating Aimogen Pro.

What OpenAI unlocks#

Features that require an OpenAI key and do not work with any other provider:

Feature	Why
AI Assistants	Uses the OpenAI Assistants API
AI Batch Requests	Uses the OpenAI Batch API (roughly 50% cheaper, asynchronous)
AI Model Training	Fine-tuning is OpenAI-only in the plugin
OpenAI Vector Stores	One of four vector store backends
Realtime voice chatbot	Uses the OpenAI Realtime API
Sora video generation	`sora-2`, `sora-2-pro`
`gpt-image-*` image generation	Alternatives exist, but this is the default
Whisper / `gpt-4o-transcribe` speech-to-text	The only transcription backend
OpenAI text-to-speech	Alternatives exist (ElevenLabs, Google, Azure)
`text-embedding-3-*` embeddings	Alternatives exist (Google, Ollama)

Text generation, the chatbot, agents, OmniBlocks and the content editor all work fine without OpenAI.

The Responses API#

Settings › Advanced AI Settings › Enable Usage Of OpenAI's Responses API switches text generation to the newer Responses endpoint.

It is switched on automatically when you select a model that only exists there, such as o1-pro.
Leave it off unless you have a reason. The classic chat completions path is the more heavily exercised code path in the plugin.

The same OpenAI / AiomaticAPI API Keys textarea serves three destinations, chosen by Main API Service Provider Selector:

openai → api.openai.com
azure → your Azure endpoint, see Microsoft Azure OpenAI
custom → your own base URL, see Custom OpenAI-compatible APIs

You cannot use OpenAI and Azure simultaneously through this slot. You can combine OpenAI with Anthropic, Google, xAI and the rest, because those have their own key fields.

Endpoints used#

https://api.openai.com/v1/chat/completions
https://api.openai.com/v1/responses
https://api.openai.com/v1/embeddings
https://api.openai.com/v1/images/generations
https://api.openai.com/v1/audio/speech
https://api.openai.com/v1/audio/transcriptions
https://api.openai.com/v1/assistants
https://api.openai.com/v1/vector_stores
https://api.openai.com/v1/batches
https://api.openai.com/v1/fine_tuning/jobs
https://api.openai.com/v1/videos

If your host uses an outbound allowlist, api.openai.com must be reachable.

Data retention#

Settings › Advanced AI Settings › Store AI Prompts On OpenAI's Part maps to the OpenAI store parameter. When off, OpenAI is asked not to retain the request for model improvement or logging beyond what its policy requires.

The same option exists separately for the chatbot (AI Chatbot › API Parameters › Store AI Prompts On OpenAI's Part) and per rule.

Turn it off if you process personal data. See Data, GDPR and retention.

Common problems#

insufficient_quota / exceeded your current quota No credit on the account, or a spend cap was hit. Check platform.openai.com/usage. A ChatGPT subscription does not help.

invalid_api_key after pasting a correct key Usually a copied newline or trailing space, or a key from a different organisation. Re-copy and use Test API key.

HTTP 429 during bulk runs Your account tier limits requests per minute. Raise Delay Between API & Scraping Requests (ms), add more keys to the textarea, or enable Reliability Mode with key rotation.

model_not_found on a model you can see in the dropdown Some models require organisation verification or a minimum spend tier before OpenAI enables them on your account. Check the model page in the OpenAI docs.

Permission error mentioning organisation Clear the OpenAI Organization ID field, or set it to the correct organisation.

OpenAI

Get a key#

Configure#

Models#

What OpenAI unlocks#

The Responses API#

Azure and AiomaticAPI share this field#

Endpoints used#

Data retention#

Common problems#

Related#