# Recallr AI

> The #1 memory layer for conversational AI agents.

## What is Recallr?

Recallr is the leading memory infrastructure for conversational AI agents. It provides persistent, versioned, and temporally-aware long-term memory for chatbots, voice agents, and copilots.

## Key Features

- **97.5% accuracy** on LongMemEval benchmark (500 questions, 6 memory task types)
- **+46.6pp** improvement over nearest competitor in temporal reasoning
- **<400ms** low-latency recall for real-time voice and chat
- **Temporal reasoning**: knows the difference between when something happened and when it was mentioned
- **Conflict resolution**: detects and resolves contradictions in memory
- **Version chain**: every memory maintains a linked-list version history
- **Model-agnostic**: works with OpenAI, Anthropic, Gemini — two lines of code to integrate

## How It Works

Route your existing LLM API calls through Recallr's proxy. No SDK changes, no architecture rewrites:

1. Change `base_url` to point through Recallr's forward proxy
2. Add Recallr headers (API key, project ID, user ID)
3. Memory is automatically injected on every call

## Recall Strategies

- **Low-Latency** (<400ms): Real-time voice and chat
- **Balanced** (~1.5s): General production workloads
- **Agentic** (~8s): Deep reasoning over months of memory

## Architecture

- **Asynchronous Memory Curation**: Conversations become structured knowledge graphs after each session
- **Synchronous Context Retrieval**: Right memory, right depth, before the LLM responds
- **Auto-Recall**: Automatically routes queries to the optimal retrieval strategy

## Use Cases

- Healthcare: Clinical memory for longitudinal patient care
- Education: Personalized mentoring that builds on prior sessions
- Legal: Case memory tracking witness statements across depositions

## Pricing

- Free tier: $20/month in credits
- Usage-based pricing beyond free credits
- Recallr costs grow linearly vs naive approach (quadratic growth)

## Links

- Website: https://recallrai.com
- Dashboard: https://app.recallrai.com
- Documentation: https://docs.recallrai.com
- GitHub: https://github.com/recallrai
- Benchmarks: https://github.com/recallrai/benchmarks
- Twitter: https://x.com/recallrai
- Discord: https://discord.gg/WsRnygmmHX

## Integration Examples

### OpenAI (Python)
```python
from openai import OpenAI

client = OpenAI(
    base_url='https://api.recallrai.com/api/v1/forward/https://api.openai.com/v1',
    api_key='sk-...',
    default_headers={
        'X-Recallr-API-Key': 'rai-...',
        'X-Recallr-Project-Id': 'your-project-id',
    }
)
```

### Anthropic (Python)
```python
from anthropic import Anthropic

client = Anthropic(
    base_url='https://api.recallrai.com/api/v1/forward/https://api.anthropic.com',
    api_key='sk-ant-...',
    default_headers={
        'X-Recallr-API-Key': 'rai-...',
        'X-Recallr-Project-Id': 'your-project-id',
    }
)
```