OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
Agent Memory & RAG

Telemem

Telemem is a Python agent-memory library that stores and retrieves information across multi-turn dialogues, with semantic deduplication and optional character-memory and video-reasoning capabilities. It can be integrated as a Mem0-style memory layer or extended with an MCP server.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Telemem — OpenViST project profile in Agent Memory & RAG.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use Telemem when an agent needs to retain conversational context beyond a single exchange, or when a project needs the additional character-memory or video pipeline. The core package is for text memory; MCP and video support are optional extras.

Who it is for

  • Developers adding persistent memory to conversational AI agents.
  • Teams building role-play, companion or multi-persona assistants that need separate character profiles.
  • Developers who want an optional video-memory and question-answering pipeline.

Documented uses

  • Save conversation details and retrieve them later through a natural-language query.
  • Reduce semantically similar memories before storage.
  • Keep character or persona information available across interactions.
  • Index video-derived frames and captions for retrieval and question answering.
Before you choose

Requirements & limits

The repository specifies the Apache-2.0 license and documents installation through pip. The README shows separate core, MCP, video and all-extras packages.

  • The default Memory() configuration inherits provider settings from mem0ai; the quick-start example sets an OpenAI API key.
  • A local Qwen and FAISS configuration is shown separately and must be loaded explicitly.
  • Video and MCP capabilities are installed as optional extras rather than being part of the core text-memory install.
  • The README describes Mem0-style imports and matching add() and search() shapes, but compatibility should be assessed against the specific APIs an application uses.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Add dialogue memory and retrieve a relevant detail

Use Telemem's documented Python example to store a conversation and search it with a question.

Before you start

  • Python environment with the Telemem core package installed.
  • OpenAI API key for the default provider configuration, or an explicitly loaded alternative configuration.

Steps

  1. Install the core package with the command below.

    Documented command
    pip install telemem

  2. Set OPENAI_API_KEY for the default provider settings, or load a configuration file with load_config and pass it to Memory.

  3. Import telemem as mem0 and create a Memory instance.

  4. Call memory.add with messages and a user_id, then call memory.search with a question and the same user_id.

  5. Read matching items from the returned results collection.

Expected outcome

The example stores a multi-turn exchange and returns matching memory text for a later question.

Guide limitations

  • The shown default configuration uses provider settings inherited from mem0ai and requires an OpenAI API key.
  • The README's local Qwen and FAISS path requires explicitly loading the repository configuration.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
9 October 2026
Original website title
TeleAI-UAGI/telemem · Official homepage title · retained in its source language
How we review sourcesSuggest a correction

Related options

Mem0 — Telemem presents itself as a Mem0-style drop-in replacement and documents the `import telemem as mem0` pattern.

Explore project →