Your first connection.
Create an API key and use it to connect your application to the models on the platform.
- One key reaches every model
- Shown once, kept on your server
From renewable power
to your API call.
Your ideas deserve a strong foundation. Discover AI APIs, our approach to infrastructure, and a commitment to renewable energy.
Coming soon
WHERE TO GO
Everything you need, one click away.
Pick the page for the job: the models, your first request, audio, the full reference or the guides.
Models
Every model this environment serves, with its capabilities, limits and prices.
Quickstart
Sign up, add credit, create a key and send your first request.
Speech to text
Send a recording, get a transcript back.
API reference
Every endpoint, parameter and response, from the published contract.
Docs
How the platform works: keys, the base URL, errors and limits.
ONE PLACE TO BEGIN
From an idea to your first API call.
One organisation for your API keys, usage and billing. A simpler place to start building with Sophea.
A clear view of usage.
See usage across your organisation, with filters for API keys, models and dates.
- Usage per key and per model
- Every call costed in EUR
Billing in one place.
Keep your prepaid balance in view, and top it up when you need to.
- Prepaid credit in EUR
- Your balance always in view
The platform is being built. Access will open once services are ready.
OPENAI-COMPATIBLE API
Change the base URL.
Keep your code.
Sophea Platform speaks the OpenAI API, so the SDKs and tools you already use work against our models. Point them at one base URL and send your first request.
- Sign up
Your first sign-in opens an organisation for you.
- Add credit
A prepaid top-up unlocks API keys.
- Create a key
Keep it on your server and call any model.
POST /chat/completions
curl "https://platform.sophea.ai/api/v1/chat/completions" \
-H "Authorization: Bearer $SOPHEA_API_KEY" \
-H "Content-Type: application/json" \
-d '{"model":"kimi-k3","messages":[{"role":"user","content":"Say hello in one sentence."}]}'ONE KEY, EVERY MODEL
Every model.
One API key.
One API key reaches every model this environment offers, through 3 public, OpenAI-compatible endpoints. Each model keeps its own configured deployment, with its capabilities and limits documented in the catalogue.
/chat/completions
Text conversations, as JSON or streamed.
/audio/transcriptions
A recording in, a transcript out.
/audio/speech
Text in, spoken audio out.
THE FOUNDATION MATTERS
Our infrastructure.
A clear direction.
The infrastructure behind an API matters as much as the interface in front of it. Our approach puts direct model access and European deployment at the centre.
- A European deployment focus: EU regions for the application and gateway.
- Dedicated GPU inference: inference on dedicated GPUs. Processing location varies by model.
- An explicit serving path: each model uses its configured deployment, without fallback to another provider.
OUR ENERGY COMMITMENT
Our commitment, supported by KIEFER SA Green Energy Projects.
Powered by (opens in a new tab)INTELLIGENCE WITH RESPONSIBILITY
Great ideas need
a future to grow in.
We believe the progress AI makes possible should go hand in hand with a more responsible energy future.
Our commitment to 100% renewable electricity for Kiefer-operated hardware is supported by KIEFER SA green energy projects. It does not cover rented GPU or cloud providers.
Renewable electricity is one part of the picture. Hardware, cooling and the wider infrastructure also have an environmental footprint.
Better choices behind the technology.
Kiefer's energy projects