• Target branch: Verify that the pull request targets the dev branch. Not targeting the dev branch will lead to immediate closure of the PR.
• Description: Provide a concise description of the changes made in this pull request down below.
• Changelog: Ensure a changelog entry following the format of Keep a Changelog is added at the bottom of the PR description.
• Documentation: If necessary, update relevant documentation Open WebUI Docs like environment variables, the tutorials, or other documentation sources.
• Dependencies: Are there any new dependencies? Have you updated the dependency versions in the documentation?
• Testing: Perform manual tests to verify the implemented fix/feature works as intended AND does not break any other functionality. Take this as an opportunity to make screenshots of the feature/fix and include it in the PR description.
• Agentic AI Code: Confirm this Pull Request is not written by any AI Agent or has at least gone through additional human review AND manual testing. If any AI Agent is the co-author of this PR, it may lead to immediate closure of the PR.
• Code review: Have you performed a self-review of your code, addressing any coding standard issues and ensuring adherence to the project's coding standards?
• Title Prefix: To clearly categorize this pull request, prefix the pull request title using one of the following:
  • perf: Performance improvement

Changelog Entry

Description

• Processes embedding requests to external APIs (OpenAI, Azure OpenAI, Ollama) in parallel instead of sequential batches.

Before:

• Sequential batch processing: Batch 1 → wait → Batch 2 → wait → Batch 3 → ... → Batch N
• For 6000 chunks with batch_size=1: 6000 sequential HTTP requests (!!!)
• Total time ≈ 6000 × request_latency (e.g., 6000 × 50ms = 5 minutes)

After:

• Parallel batch processing: All batches sent simultaneously via asyncio.gather()
• For 6000 chunks with batch_size=1: All 6000 requests execute in parallel
• Total time ≈ 1 × request_latency + overhead (e.g., <10 seconds for 6000 chunks)
• ~30-50x speed improvement observed in production

Added

• Async embedding generation functions:
  • generate_openai_batch_embeddings_async() using aiohttp
  • generate_azure_openai_batch_embeddings_async() using aiohttp
  • generate_ollama_batch_embeddings_async() using aiohttp
• generate_multiple_async() function for parallel batch coordination
• Import of aiohttp and asyncio modules

Changed

• get_embedding_function() now returns synchronous wrapper around async implementation
  • External API remains unchanged: still returns synchronous callable
  • Internal implementation uses asyncio.run() to execute parallel requests
• generate_embeddings() now uses asyncio.run() for async execution
• All batch embedding functions converted from synchronous requests.post() to async aiohttp.ClientSession()

Fixed

• Massive performance bottleneck when processing large documents with external embedding APIs
• Sequential processing causing unnecessary delays for models that only support batch_size=1

Breaking Changes

• NO BREAKING CHANGES:
  • All callers remain synchronous and unmodified
  • API contract is identical: embedding_function(texts, prefix, user) still returns embeddings synchronously
  • Uses "sync wrapper around async code" pattern via asyncio.run()
  • No changes required to calling code in retrieval.py or elsewhere

Contributor License Agreement

By submitting this pull request, I confirm that I have read and fully agree to the Contributor License Agreement (CLA), and I am providing my contributions under its terms.