FAQ

What does this project do in one sentence?

It removes boilerplate for running OpenAI Batch jobs in Python, from building JSONL inputs to downloading and parsing outputs.

Does it stream responses?

No. The Batch API is asynchronous and writes results to an output JSONL file when processing completes.

Which endpoints are supported?

The helper focuses on /v1/chat/completions and /v1/embeddings because they are reliable in batch mode. Ensure your per-line url matches the batch endpoint you pass during creation.

How do I see progress while waiting?

Use the logging-based progress callback:

import logging
from openai_batch_helper import status_progress_logger

logging.basicConfig(level=logging.INFO)
job.wait_for_completion(on_update=status_progress_logger(heartbeat_seconds=30))

What errors can be raised?

• EmptyBatchError: You attempted to submit an empty JSONL file.

• BatchNotCompletedError: You tried to download results before completion.

What Python versions are supported?

Python 3.9+.

Can I customize result parsing?

Yes, pass an extractor to map_by_custom_id to return any value you want for each output row.