Pre-launchTallrig is being built in the open and is not yet open for production customers. General availability: late 2026 – 2027.

Colour theme

02 / 06 — Network

Your edge, end to end

DNS, runtime, firewall, and delivery — on infrastructure we own.

Everything between your users and your code runs through Tallrig's own reverse proxy and gateway, on a cloud server we rent, in one region today. No third-party CDN in the request path. Our WAF and DDoS shield protect the Tallrig API today, and putting them in front of your own app is coming soon; your app's traffic already passes a per-IP rate limit at our gateway.

Request access Read the docs

V8Isolates at the edge, in preview

OwnReverse proxy and gateway, no third-party CDN

OursSoftware stack, end to end

Live — running in production · Preview — built, but not yet proven for customers, and parts may not work yet · Coming soon — on the roadmap

01 / 08

Edge runtime

Preview

V8 isolates for latency-sensitive code, with no container to boot. In preview: edge functions are not running for customers yet.

The runtime speaks standard Web APIs — Request, Response, fetch, and streams — so the same code runs at the edge and in your tooling with nothing proprietary to learn. Because isolates share a process instead of booting a container, there's no scheduler queue between a request and your code, and the same deploy serves every request from the moment it goes live.

Isolates, not containers — no OS image to boot per requestStandard Web APIs — no proprietary runtime to learnOne deploy, one place to look — the same code serves every request

02 / 08

DNS hosting

Live

Authoritative DNS served from our own nameservers, with a clean API.

Authoritative DNS is managed for you the moment a domain is attached — apex and wildcard records included — and edits take effect on our nameservers in seconds rather than waiting out a long TTL. It's the same DNS that points the rest of the platform at our own boxes, so there's no separate provider to wire up or keep in sync.

Apex and wildcard recordsChanges propagate in seconds, not hoursManaged automatically when you attach a domain

03 / 08

Object storage

Preview

An S3-compatible object storage API, backed today by our cloud provider's object storage.

A drop-in S3 API means existing SDKs and tools work unchanged, with signed URLs and fine-grained access control for the objects you don't want public. The service is in preview while we bring it up on our current infrastructure, and no egress pricing is promised yet.

Drop-in S3 API compatibilitySigned URLs and access control

04 / 08

AI gateway

Preview

One endpoint for every model — cached, routed, rate-limited, and observable.

One endpoint and one key sit in front of every model — our own GPUs and external providers — with response caching for repeated prompts, automatic failover when a provider degrades, and per-key budgets that stop a runaway loop cold. Every call is traced end to end, so cost and latency are something you can inspect, not guess at.

Unified API across model providers and our own GPUsResponse caching and automatic failoverPer-key budgets and full request tracing

05 / 08

WAF & rate limits

Coming soon

Application firewall and rate limits on our own servers. The firewall protects the Tallrig API today; in front of your apps is coming soon.

Today the firewall, DDoS absorption and per-route rate limits protect the Tallrig API itself: an OWASP rule set that blocks by default, on our own capacity. Your apps already get a per-IP rate limit at our gateway; the firewall and per-route limits in front of them, with your own rules on top, are coming soon.

Managed OWASP rule set, blocking by defaultPer-IP rate limits at the gateway today; per-route limits coming soonDDoS absorption on our own capacity

06 / 08

Reverse tunnel

Preview

Expose a local or private service through our edge without port-forwarding.

A secure, outbound-only tunnel exposes a local or private service through our edge with a stable public URL — no port-forwarding, no firewall holes, no inbound exposure. It's built for demos, webhook development, and hybrid setups where part of your stack lives somewhere we don't.

Secure outbound-only tunnelsStable public URLs for private servicesGreat for demos, webhooks, and hybrid setups

07 / 08

Video pipeline

Coming soon

Ingest, transcode, and stream adaptive video from one pipeline.

Ingest once and the pipeline transcodes to multiple bitrates automatically, then delivers adaptive HLS or DASH from the edge so playback holds up across connections. Thumbnails and per-asset analytics come with it, so you don't bolt a second service onto the side just to see what's being watched.

Automatic multi-bitrate transcodingAdaptive HLS/DASH delivery from the edgeThumbnails and per-asset analytics

08 / 08

Privacy analytics

Preview

Cookieless, first-party traffic insight that respects your users.

Traffic insight without cookies or cross-site tracking means there's no consent banner to bolt on and no third party quietly profiling your users. The data is first-party and yours, collected by a lightweight script that won't drag down the page it measures.

No cookies, no cross-site trackingFirst-party data you actually ownLightweight script, no consent banner needed

Network — frequently asked questions

Is there a third party in my request path?

No. Your traffic runs only through Tallrig's own edge, on servers we operate — we don't proxy it through anyone else's.

How fast are edge cold starts?

We have not published a cold-start measurement yet, so we make no latency promise. The edge runtime runs V8 isolates, so there's no container to spin up before your code executes. It is in preview and not yet running customer functions.

Is the object storage really S3-compatible?

Yes — it's a drop-in S3 API, with signed URLs and fine-grained access control, and it is in preview while we bring it up on our current infrastructure.

Do I have to manage DNS separately?

No. When you attach a domain its authoritative DNS is managed for you — apex and wildcard records — and changes take effect on our nameservers in seconds.

What does the AI gateway do?

It gives you one endpoint for every model — our own GPUs and external providers — with response caching, automatic failover, per-key budgets, and full request tracing.

Build on Network today

Signups are invite-only right now — request access and we'll follow up.

Request access Read the docs