openai

Use curl to send JSON POST requests to https://api.openai.com/v1/chat/completions with your OPENAI_API_KEY in the Authorization header, specifying the model (GPT-4o, GPT-4, or GPT-3.5), messages array, and content-type as application/json. This Skill provides ready-to-use curl commands for chat completions.

Yes. OpenAI's DALL-E 3 endpoint generates images via curl POST requests, and Whisper transcribes audio files through the /audio/transcriptions endpoint. Both require your API key, proper headers, and adherence to rate limits and payload format conventions.

Send curl POST requests to OpenAI's /embeddings endpoint with your text and model choice, receiving vectors for semantic search and RAG pipelines. This Skill handles the JSON formatting, authentication, and response parsing needed to integrate embeddings into automation workflows.

You need a valid OPENAI_API_KEY, curl-enabled environment, access to https://api.openai.com/v1, and knowledge of JSON payload structure and content-type headers. This Skill removes the formatting overhead with pre-built scripts.

Streaming requires setting the stream parameter to true in your JSON payload; curl will output server-sent events. Rate limits are enforced per endpoint; respect response headers and implement backoff logic. This Skill documents both patterns in its references.

Yes. Combine GPT chat completions with embeddings for retrieval-augmented generation (RAG): query embeddings find relevant context, send it with your question to GPT, and format the response. This Skill provides curl examples for each step in the workflow.