We’re back with a powerful new capability in DocWire SDK. Version 2025.07.14 adds support for OpenAI embeddings, modularizes the HTTP client, and improves MIME-based content handling across AI components.
Full release notes: https://github.com/docwire/docwire/releases/tag/2025.07.14
New Features
1 · OpenAI Embeddings Support
- New
openai::embedchain element enables developers to generate semantic embeddings from text using OpenAI’s latest models:-
text-embedding-3-small -
text-embedding-3-large -
text-embedding-ada-002
-
- Includes CLI support (
--openai-embed,--openai-embed-model) and full documentation with examples.
2 · MIME to Extension Utility
- New
content_type::by_file_extension::to_extensionutility maps MIME types back to standard file extensions.
Improvements
More Robust AI Input Handling
AI components now detect input types using MIME (e.g.,text/plain,image/png) instead of relying solely on file extensions. This improves accuracy in content classification and avoids misfires.Smarter HTTP Uploads
Thehttp::Postchain element can now infer file extensions from MIME types automatically, reducing friction in stream-based uploads.
Refactor
-
HTTP Client Modularization
http::Postis now part of a dedicateddocwire_httplibrary, improving modularity and simplifying future expansion.
Fixes
-
Windows Build Fix
Resolved vcpkg failure on Windows by explicitly linking
libxml2. - README Fixed a formatting issue in the documentation.
A whisper of meaning, in numbers cast
New dimensions open, built to last
For search and reason, a powerful key
Unlocking knowledge for all to see
Try It Now
- GitHub: https://github.com/docwire/docwire
- Release: https://github.com/docwire/docwire/releases/tag/2025.07.14
- Sourceforge https://sourceforge.net/projects/docwire/files/2025.07.14/
This release brings DocWire one step closer to being the go-to C++ SDK for hybrid AI-powered backends and document pipelines.
We welcome your feedback, issues, or questions.
— The DocWire Team
Top comments (0)