# Release Notes for v0.4.0

## Highlights

- Support [Evaluation (beta)](../evaluation.mdx) tool to evaluate the performance and reliability of the Chat Engine’s outputs.
  - Current support key metrics:
    - Factual Correctness
    - Semantic Similarity
- Support new LLM providers
  - [Gitee AI](../llm.mdx#gitee-ai)
  - Test new OpenAI-like providers
    - [vLLM](../llm.mdx#vllm)
    - [Xinference](../llm.mdx#xinference)
- Support new embedding model providers
  - [Gitee AI](../embedding-model.mdx#gitee-ai)
  - [Amazon Bedrock](../embedding-model.mdx#amazon-bedrock)

## Improvements

- Limit the upload file size via `max_upload_file_size` (10MB by default) parameter on site setting
- Support download the reference file in the chat page

If you are deploying Autoflow using docker, please follow the [Upgrade](../deploy-with-docker.mdx#upgrade) guide to upgrade your Autoflow.