Stop guessing which one. Put them side by side.

The grid says what these tools are. This says what builders did about it — one decision per person, changeable whenever you change your mind.

# Tool comparison — Lindy vs pg-boss vs Activepieces

Source: Kbaise, a directory of tools that work with Lovable projects.
Pick one and tell me why before writing any integration code.

## Lindy (lindy)
- URL: https://www.lindy.ai
- Category: Automation & background jobs
- Cost: paid — No standing free plan; 7-day trial via Slack invite. Plus $29.99/user/mo for 3,000 credits, Pro $99.99 for 15,000, Max $199.99 for 35,000. Enterprise custom.
- What it is: You describe an assistant in plain English and Lindy builds the automation: email triage, meeting notes, CRM updates, outbound phone calls. MCP support on every tier connects arbitrary tools. Credits scale with task depth, from 2-250 for everyday tasks up to 2,500 for big builds.
- Why builders pick it: The least technical entry point on this list: no canvas, no nodes, just a conversation. Reach for it when the job is an ongoing assistant watching an inbox or calendar rather than a deterministic pipeline.
- Trap: Billing is per seat and per credit, and credit costs are hard to predict: one 'big build' can consume 2,500 credits, most of a Plus seat's monthly 3,000 allowance.
- Full dossier: /api/public/tools/lindy

## pg-boss (pg-boss)
- URL: https://github.com/timgit/pg-boss
- Category: Automation & background jobs
- Cost: free — Free and MIT-licensed, no hosted tier and no per-job billing. Your only cost is whatever your Postgres instance already costs.
- What it is: A Node.js job queue that uses Postgres SKIP LOCKED for concurrency instead of Redis. Provides cron-scheduled jobs, deferred and throttled jobs, retries with exponential backoff, job dependencies via completion callbacks, and automatic archiving and pruning of finished jobs.
- Why builders pick it: The zero-new-infrastructure option: if you have Postgres you already have a queue, and jobs can commit in the same transaction as your data, which kills the dual-write bug where the row saves but the job vanishes.
- Trap: Needs a long-running Node process, so it does not work on serverless. Throughput ceiling is your database: fine for thousands of jobs a minute, not hundreds of thousands.
- Wiring: npm install pg-boss
- Full dossier: /api/public/tools/pg-boss

## Activepieces (activepieces)
- URL: https://www.activepieces.com
- Category: Automation & background jobs
- Cost: mixed — Community edition free under MIT with no cap on runs, users or flows. Cloud Free: daily refreshing credits. Plus $16/mo (up to 5 users), Team $166/mo (25 users). Overage $0.007/credit.
- What it is: Open-source Zapier alternative under a true MIT licence. One credit per flow run regardless of how many steps it contains, which makes long workflows cheap. Ships Flows, Agents, Chat, Tables and MCP servers on every tier, and pieces are TypeScript so you can write your own.
- Why builders pick it: The pick when licence purity matters: MIT lets you embed and resell it, which n8n's fair-code licence forbids. Per-run rather than per-step billing rewards big workflows, and self-hosting has no artificial run cap.
- Trap: The open-source build excludes Agents, Chat, Projects, API access, SSO and audit logs, so 'MIT licensed' does not mean feature parity with cloud. AI model steps cost 2 to 20 credits unless you bring your own key.
- Wiring: docker run -p 8080:80 activepieces/activepieces
- Full dossier: /api/public/tools/activepieces

## Quick read

- Free with no card: pg-boss.

## Rules
1. Prefer the free tier when no budget was stated, and say what the ceiling is.
2. Read the full dossier before integrating.
3. Fetch /api/public/models before writing any AI model ID.

Agents can fetch the same thing: GET /api/public/compare?slugs=lindy,pg-boss,activepieces