Overview
The OpenAI connector allows you to integrate directly with OpenAI's REST API using an API key for authentication. This provides access to a wide range of artificial intelligence models and tools, including GPT for advanced chat completions, DALL-E for image generation, and Whisper for audio transcription.
By configuring the connector, you can make direct calls to OpenAI API endpoints. This enables you to build powerful workflows that can analyze text, generate content, moderate user input, create text embeddings for semantic search, and interact with the Assistants API to build complex, stateful conversational agents. The connector supports handling various objects like Models, Assistants, Threads, Messages, and Files, giving you granular control over your AI-powered integrations.
Capability summary
Authentication
Auth method: API Key
Required scopes
]
Supported objects
| Object | Read | Write | Notes |
|---|---|---|---|
| Model | Yes | No | Represents an AI model, such as GPT-4 or DALL-E 3. |
| Assistant | Yes | Yes | A purpose-built AI that uses models and tools to respond to user queries. |
| Thread | Yes | Yes | Represents a conversation session between an Assistant and a user. |
| Message | Yes | Yes | A message created by a user or an Assistant within a Thread. |
| File | Yes | Yes | A document or data file uploaded for processing by models or tools. |
| Fine-Tuning Job | Yes | Yes | A job to train a model on a custom dataset. |
| Vector Store | Yes | Yes | A collection of embedded files for retrieval, used with the Assistants API. |
Supported actions
| Action | Description |
|---|---|
| create_chat_completion | Generates a model response for a given chat conversation or prompt (GPT models). |
| generate_image_dall_e | Creates an image from a text prompt using a DALL-E model. |
| create_text_embedding | Returns a vector representation of a given input text string. |
| transcription_whisper | Transcribes audio into text using the Whisper model. |
| analyze_image_vision | Analyzes an image and responds to questions about its content (GPT-4 with Vision). |
| moderation_check | Checks if a text string violates OpenAI's content policy. |
| message_assistant | Adds a message to an Assistant Thread and triggers a run to get a response. |
| text_translation | Translates text from one language to another using a GPT model. |
Setup steps
- Log in to OpenAI Platform: Access your OpenAI account at platform.openai.com.
- Navigate to API Keys: In the left-hand navigation menu, select the "API Keys" section.
- Create a New Secret Key: Click the "Create new secret key" button. Give your key a descriptive name (e.g., "Flows360 Production") for easy identification.
- Copy the API Key: Your new API key will be displayed once. Copy it immediately and store it securely, as you will not be able to view it again.
- Enter Key in Flows360: In the Flows360 OpenAI connector configuration, paste the copied key into the "API Key" field.
- Configure Authorization Header: Ensure the request header is set to
Authorization: Bearer YOUR_API_KEY, whereYOUR_API_KEYis the token from the previous step.
Limitations
The connector operates within the rate limits and quota restrictions of your OpenAI API account tier. Check your OpenAI account for specific limits on tokens per minute (TPM), requests per minute (RPM), and monthly usage quotas. Exceeding these limits will result in API errors (HTTP 429). The connector does not manage model fine-tuning processes directly; it can only initiate and monitor jobs. File and Vector Store management may require multi-step workflows for uploads, processing, and attachment to objects like Assistants or Threads.