Comparison
| Feature | Svix | Hookdeck | Convoy |
|---|---|---|---|
| Free tier | 25k msg/mo | 25k evt/mo | Open source |
| Managed / self-host | Managed | Managed | Both |
| Dashboard UI | ✅ | ✅ | ✅ |
| Portal for customers | ✅ | ❌ | ❌ |
| Best for | SaaS platform | Dev events | Self-host |
Svix, Hookdeck and Convoy compared. Reliable webhook delivery without building your own queue.
| Feature | Svix | Hookdeck | Convoy |
|---|---|---|---|
| Free tier | 25k msg/mo | 25k evt/mo | Open source |
| Managed / self-host | Managed | Managed | Both |
| Dashboard UI | ✅ | ✅ | ✅ |
| Portal for customers | ✅ | ❌ | ❌ |
| Best for | SaaS platform | Dev events | Self-host |
This calculator estimates the monthly cost of outsourcing webhook delivery to a managed service. You enter your monthly event volume, the number of endpoints or consumers receiving those events, how many retry attempts you need, and your team size, and it compares the resulting bills for Svix, Hookdeck and Convoy. Webhook platforms meter differently — some price on events published, others on total delivery attempts (a single event fanned out to five endpoints with retries can count as many billable operations) — so the cheapest provider can flip as fan-out and retry policy change. Modeling this before you commit avoids the common surprise of an invoice that scales with deliveries rather than with the events your application actually produces.
It depends on the provider: some meter unique events published to the platform, while others meter each outbound delivery attempt. With fan-out to multiple endpoints and automatic retries, delivery-attempt pricing can cost several times more than the event count suggests, so check which unit your plan uses.
In-house delivery is feasible, but you have to build and operate retry queues with exponential backoff, signature verification, endpoint health tracking, replay, and delivery logs. At low volumes a managed service is usually cheaper than the engineering time; at very high volumes, self-hosting an open-source option such as Convoy on your own infrastructure can cost less than per-event pricing.
Retries multiply the work the platform performs for a single event. If a consumer endpoint is down and each event is retried several times, your billable attempts rise proportionally — which is why a small number of persistently failing endpoints can dominate costs on attempt-metered plans.