Built for production scale.

No. Your monthly commitment reserves guaranteed network capacity, like a dedicated lane: the capacity exists whether you use it or not, which is exactly what lets us guarantee throughput at peak. If you consistently use less, downgrade to a smaller tier; if you’re bursty, talk to us and we’ll size it right.

Nothing breaks. Jobs keep running and the extra volume is billed at your tier’s overage rate. You’ll get a usage alert at 80% and 100% so it’s never a surprise.

Upgrades apply immediately, the higher capacity kicks in right away, prorated. Downgrades take effect at your next billing cycle.

No annual lock-in. Plans are monthly and you can cancel anytime. Annual commitments with discounted rates are available for enterprise volumes.

Total bytes delivered to your storage, that’s the only meter. Pay as you go is metered per megabyte, so you are billed for what you actually received. Failed jobs are never billed, and we don’t charge for bandwidth, requests, retries, or webhooks.

No. Pay as you go is billed per megabyte: 1.4 GB delivered is billed as 1,400 MB, never as 2 GB. Providers that price per job or per started gigabyte make you pay for bytes they never delivered.

Yes, 100 GB free, once per account, with no expiry date. A card is required at signup so the usage past the free 100 GB can be charged. Five minutes from signup to your first delivered file.

YouTube and Spotify at full scale. Need a source we don’t list? Ask, adding platforms is a roadmap conversation, not a no.

Video (MP4, MKV) up to source quality, audio-only extraction, and metadata. Tell us your pipeline’s expected input and we’ll match it.

You include your bucket credentials in the request. Tornado downloads the source, processes it on our workers, and streams the result directly to your bucket. Your file never touches our long-term storage.

AWS S3, Azure Blob, Google Cloud Storage, Cloudflare R2, Alibaba OSS, and any S3-compatible endpoint (MinIO, DigitalOcean Spaces, Backblaze B2, Wasabi).

First file lands in your bucket within minutes of the API call. Sustained throughput scales with your tier, our measured network peak is 38.4 Gbps.

Yes, that’s our managed delivery service: we’ve shipped a ~100,000-video research dataset (~32 TB) straight into a customer’s S3, reconciled and verified. If you have a large ingestion project, book a call and we’ll scope it.

Below 0.002%, measured across the whole pipeline. Failed jobs auto-retry and webhook with the failure reason, and failures are never billed.

Every session runs on a dedicated network we own and operate ourselves, not rented proxies. An automatic feedback loop reroutes around blocks in real time, so they never reach your jobs. Full architecture details are shared on founder calls.

Contractual uptime SLAs per tier: 99.5% on Starter, 99.7% on Medium, 99.9% on Advanced, each backed by a money-back guarantee.

Tornado API is transport infrastructure: we move publicly accessible content from A to B, the way a CDN or ISP carries traffic. What you do with the content, and the rights required for your use case, is your responsibility. Most of our customers run legal-reviewed workflows (AI training under licensing agreements, analytics, creator-authorized repurposing).

No. Tornado only handles publicly accessible URLs. Customers are responsible for ensuring their use complies with applicable laws and platform terms.

Delivered files stream to your bucket and never touch our long-term storage. A short-lived cache may hold frequently requested items to speed up repeat deliveries, then expires automatically.

Yes, we’re a French company operating under EU law. A DPA is available on request.

Yes. Card self-serve for standard tiers; invoicing (net 30) and wire transfer for enterprise volumes.

Email support on all plans, shared Slack channel and priority response on enterprise tiers. You’ll talk to the person who built the system, not a ticket queue.

Anytime, 30 minutes with the founder, no sales script. Come with your hardest technical question.