One API for every model
Call language, vision and audio models with the same request shape. Switch models by changing one line, not your stack.
one API.
every model.
Access and deploy frontier language, vision and audio models through a single API, on infrastructure that keeps your data where you decide.
Launching from our first AI factory at Tinapa, Calabar. Waitlist members are onboarded first, in order.
Frontier models without the infrastructure work.
Pay for what you use
Usage-based pricing with no hardware to buy or clusters to manage. Start small and scale when your product does.
Bring your own models
Serve open models as they are, or the ones you trained with Protocols.
Your data stays yours
Run inference in-region, on infrastructure built in Africa. Your prompts and outputs stay under your control.
Language, vision and audio. Same request.
Language
Chat, summarise, extract and reason over long documents in the languages your customers speak.
curl $HIPERLINKS_URL/v1/chat \-H "Authorization: Bearer $HIPERLINKS_KEY" \-d '{"model": "language-large","messages": [{ "role": "user","content": "Summarise this contract" }]}'
From waitlist to production.
01
Join the waitlist
Reserve early access. We onboard teams in waitlist order as capacity comes online.
02
Make a request
Point your code at Hiperlinks and call any model with one request shape.
03
Deploy and scale
Move to dedicated, in-region capacity when you need guaranteed throughput.
start building.
Join the waitlist and get your API key as soon as capacity comes online.