AI assistants and search decision
Wabery
A competent developer can build a narrow substitute (message routing, signed webhooks, basic Flow handling, and running simple TypeScript handlers) in weeks, but reproducing the full hosted runtime, MCP integrations, Flow encryption, team inbox, and production-grade delivery/retry guarantees is substantial and justifies using the SaaS for full product needs.
Visit website↗No open-source build does this yet
Nothing published replaces this one, so a replacement starts from an empty file. Here is what it would have to cover.
What a replacement has to do
- Accept WhatsApp inbound messages, deliver them to an AI or webhook, send replies and structured Flow submissions back to the user.
What it still won’t have
- Hosted functions runtime with invocation logs and built-in deployment tooling
- Prebuilt WhatsApp Flow encryption handshake and Flow JSON handling
- Signed, typed event delivery with built-in retries and endpoint health tracking
- CLI and MCP server integrations that let agents drive configuration from third-party LLMs
- Team inbox, roles, and UI-managed channel sandbox
What remains hard
- Product polish and ongoing maintenance
First-year cost
Keep paying
Paying is—cheaper in year one.
On cash alone, building overtakes the subscription at 3 seats.
Money you would actually spend
Time you would spend
—
What you would spend
What we assumed
The verdict above measures whether you could build it. This one is only about money.
Runnable build prompt
Build a minimal WhatsApp AI-agent platform in TypeScript: Node.js + Express for webhooks, Postgres for threads/contacts, a small worker for retries (BullMQ), and a tiny serverless-like runtime using V8 isolates or use a sandboxed Function-as-a-Service (e.g. Cloud Run) for user TypeScript handlers. In scope: (1) register and manage a WhatsApp Business channel via the Meta Cloud API, (2) receive and verify signed inbound webhooks, persist threads/contacts and deliver typed events to a webhook endpoint, (3) send outbound messages and track delivery status with retry logic, (4) execute user TypeScript handlers for simple tool calls and expose logs, (5) implement a Flow authoring JSON format and validate submissions (encryption handshake may be stubbed if Meta keys unavailable). Out of scope: visual dashboard UI, multi-tenant team inbox with roles, deep MCP/LLM integrations, and paid carrier billing automation. Provide error handling, retries, end-to-end tests for webhook signing/verification, and CI deployment scripts (Docker + one cloud provider).
How we checked
How the score was reached
- Partly verdict base52
- 3 cited sources+3
- Price verified on pricing page+3
- 3/3 assessment runs agreed+4
- Evidence score62
The base comes from the verdict. Everything under it is a check that either happened or did not, and each one is a fact frozen in this record rather than a judgement made at render time — so the same evidence always produces the same number.
How scoring works →Cited sources · 3
Every page the run actually retrieved.
- official productWabery product homepage
- official docsWabery docs (capabilities & quickstart)
- official pricingWabery WhatsApp pricing
Integrity checks
What held up, and what did not.
