Metadata-Version: 2.4
Name: unibaseio-membase-core
Version: 0.3.1
Summary: The Membase memory engine (compiled). Use it through unibaseio-membase.
License: Proprietary
Project-URL: Homepage, https://membase.ai
Project-URL: Documentation, https://docs.membase.ai
Project-URL: SDK, https://pypi.org/project/unibaseio-membase/
Requires-Python: <3.14,>=3.12
Description-Content-Type: text/markdown
License-File: LICENSE
Requires-Dist: openai>=1.50
Requires-Dist: faiss-cpu>=1.8
Requires-Dist: numpy>=1.26
Requires-Dist: pydantic>=2.7
Requires-Dist: tqdm>=4.66
Requires-Dist: tenacity>=8.5
Requires-Dist: sentence-transformers>=3.0
Requires-Dist: asgiref>=3.8
Requires-Dist: tiktoken>=0.7
Provides-Extra: multimodal
Requires-Dist: pillow>=10.0; extra == "multimodal"
Requires-Dist: httpx>=0.27; extra == "multimodal"
Requires-Dist: beautifulsoup4>=4.12; extra == "multimodal"
Provides-Extra: anthropic
Requires-Dist: anthropic>=0.40; extra == "anthropic"
Provides-Extra: ollama
Requires-Dist: ollama>=0.4; extra == "ollama"
Provides-Extra: all
Requires-Dist: pillow>=10.0; extra == "all"
Requires-Dist: httpx>=0.27; extra == "all"
Requires-Dist: beautifulsoup4>=4.12; extra == "all"
Requires-Dist: anthropic>=0.40; extra == "all"
Requires-Dist: ollama>=0.4; extra == "all"
Dynamic: license-file

# unibaseio-membase-core

The engine of [Membase](https://membase.ai), the memory layer for AI agents: conversations become
dated episodes, documents a topic tree, and search is a multi-round retrieval. 93.1% on LoCoMo,
92.6% on LongMemEval_S and 92.2% on DMR ([method and results](https://membase.ai/research)).

Compiled wheels for CPython 3.12 and 3.13 on Linux, macOS and Windows. Applications use it through
[unibaseio-membase](https://pypi.org/project/unibaseio-membase/): `pip install 'unibaseio-membase[local]'`.

Proprietary (see LICENSE); the bundled `membase_algo` is Apache-2.0 (see NOTICE).
