Backfire compared with an inbound tunnel

When each one wins

Use a tunnel when a remote service must reach a server you run locally. Use Backfire when your code must reach a remote service with a key, a guard and a log.

The same call, side by side

# inbound — a tunnel publishes a port on your machine
ngrok http 3000            # https://ab12.ngrok.app → localhost:3000

# outbound — the relay calls the target on your behalf
curl -sS https://backfireapi.base44.app/functions/forward \
  -H "Authorization: Bearer $BACKFIRE_KEY" \
  -H "Content-Type: application/json" \
  -d '{"url":"https://api.example.com/v1/items","method":"GET","timeout_ms":8000}'

Worth knowing

• Neither tool is a substitute for the other. A team can reasonably use a tunnel to receive webhooks locally and Backfire to make outbound calls in production.
• Backfire never opens a port on your machine and does not require an agent process.

Limits that stay your problem either way

• Neither tool substitutes for the other: a tunnel receives traffic, the relay sends it.
• Inbound webhooks still need a public endpoint of your own, or a tunnel — the relay cannot accept them.
• A relayed call originates from relay infrastructure, so allow-lists at the target see the relay, not you.
• Every relayed call is attributable to your key and appears in your activity log; there is no anonymity here.