⚡ dakera-rs

Rust client for Dakera AI — store, recall, and search agent memories against a Dakera instance.

Part of Dakera AI — the memory engine for AI agents.

The Dakera memory engine scores 87.8% on LoCoMo (1,540 questions, standard eval) — benchmark details

Install

# Cargo.toml
[dependencies]
dakera-client = "0.11"
tokio = { version = "1", features = ["full"] }

Quick Start

use dakera_client::{DakeraClient, Config, UpsertRequest};

#[tokio::main]
async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let client = DakeraClient::new(Config {
        base_url: "http://localhost:3300".to_string(),
        api_key: "your-key".to_string(),
        ..Default::default()
    });

    // Store a vector
    client.vectors().upsert(UpsertRequest {
        id: "vec-001".to_string(),
        values: vec![0.1, 0.2, 0.3],
        metadata: None,
    }).await?;

    // Search
    let results = client.fulltext().search("completed task", 5).await?;
    for r in results.results {
        println!("{} {:.3}", r.id, r.score);
    }

    Ok(())
}

Connect to Dakera

// Self-hosted
let client = DakeraClient::new(Config {
    base_url: "http://your-server:3300".to_string(),
    api_key: "your-key".to_string(),
    ..Default::default()
});

// Cloud (early access)
let client = DakeraClient::new(Config {
    base_url: "https://api.dakera.ai".to_string(),
    api_key: "your-key".to_string(),
    ..Default::default()
});

Documentation

→ Full docs
→ API reference
→ Rust SDK reference

Related

Part of the Dakera AI open core. The engine is proprietary. The tools are yours.