Fastpaca Context Store

Context budgeting and compaction for LLM apps. Keep long conversations fast and affordable.

• Set token budgets. Conversations stay within bounds.
• You control the accuracy/cost tradeoff.

                      ╔═ fastpaca ════════════════════════╗
╔══════════╗          ║                                   ║░    ╔═optional═╗          
║          ║░         ║  ┏━━━━━━━━━━━┓     ┏━━━━━━━━━━━┓  ║░    ║          ║░         
║  client  ║░───API──▶║  ┃  Message  ┃────▶┃  Context  ┃  ║░ ──▶║ postgres ║░
║          ║░         ║  ┃  History  ┃     ┃  Policy   ┃  ║░    ║          ║░         
╚══════════╝░         ║  ┗━━━━━━━━━━━┛     ┗━━━━━━━━━━━┛  ║░    ╚══════════╝░         
 ░░░░░░░░░░░░         ║                                   ║░     ░░░░░░░░░░░░         
                      ╚═══════════════════════════════════╝░                      
                       ░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░

Enforces a per-conversation token budget before requests hit your LLM.

• Quick start
• Getting started
• How it Works
• Policies
• Self-hosting
• Storage & Audit
• API Reference

Long conversations get expensive and slow

• More messages = more tokens = higher cost
• Larger context = slower responses
• Eventually you hit the model's limit

What Fastpaca Context Store does

Enforces per-conversation token budgets with deterministic compaction.

• Keep full history for users
• Compact context for the model
• Choose your policy (last_n, skip_parts, manual)

Quick Start

const fastpaca = createClient({ baseUrl: 'http://localhost:4000/v1' });
const ctx = await fastpaca.context('demo', { budget: 1_000_000 });
await ctx.append({ role: 'user', parts: [{ type: 'text', text: 'Hi' }] });
const { messages } = await ctx.context();

Background

We kept rebuilding the same Redis + Postgres + pub/sub stack to manage conversation state and compaction. It was messy, hard to scale, and expensive to tune. Fastpaca Context Store turns that pattern into a single service you can drop in.