One API. Zero data leaving Europe.
Polarise Drive is the fully managed inference platform for European developers and enterprises: OpenAI-compatible, consumption-priced.
- Latest Models
- Own Datacenters
- No Cloud Act Exposure
Fully managed AI inference
Drive sits on top of Polarise's own stack: GPUs, platform and datacenter are ours. OpenAI-compatible API to frontier open-source models on sovereign EU infrastructure.
Supported models
Access the latest open-source models through a single API.
Always on the latest model
Every model family has a stable :latest alias. When we upgrade a family — say from GLM 5.2 to GLM 5.3 — the alias automatically bumps to the new version, and the previous one is deprecated.
Works with any stack
The tools you already use connect with a base URL and an API key. Same clients, same code, sovereign European models behind them.
*Every client that speaks the OpenAI API.
{
"$schema": "https://opencode.ai/config.json",
"model": "polarise/glm-5.3",
"provider": {
"polarise": {
"npm": "@ai-sdk/openai-compatible",
"name": "Polarise Drive",
"options": {
"baseURL": "https://api.model-gateway.polarise.app/v1",
"apiKey": "{env:POLARISE_API_KEY}"
},
"models": {
"glm-5.3": { "name": "GLM 5.3" }
}
}
}
}
import os
from langchain_openai import ChatOpenAI
llm = ChatOpenAI(
model="glm-5.3",
base_url="https://api.model-gateway.polarise.app/v1",
api_key=os.environ["POLARISE_API_KEY"],
)
{
"agents": {
"defaults": {
"model": { "primary": "polarise/glm-5.3" }
}
},
"models": {
"mode": "merge",
"providers": {
"polarise": {
"baseUrl": "https://api.model-gateway.polarise.app/v1",
"apiKey": "${POLARISE_API_KEY}",
"api": "openai-completions",
"models": [{ "id": "glm-5.3", "name": "GLM 5.3 (Drive)" }]
}
}
}
}
model:
default: glm-5.3
provider: custom
base_url: https://api.model-gateway.polarise.app/v1
api_key: ${POLARISE_API_KEY}This is Drive
- Rise built on Drive
- Drive built on Core
- Core built on bare metal, for managed GPU rental
- Bare metal the servers we operate
- Colocation space, power & cooling for your hardware
- Datacenter infrastructure built & owned by us
- Real estate rented or bought by us
Setup in minutes
Create an API key in the portal, set your limits, and you are ready to call the API within minutes.
Create API Key
Quick create · Budget caps · Rate limits & quotas · Model restrictions · Fine-grained permissions
The Drive roadmap
The Model Gateway is production-ready today. Everything after it ships incrementally on the same OpenAI-compatible API: no go-live dates, no rewrites, no lock-in.
One platform, growing with you
OnlineA single OpenAI-compatible endpoint for every hosted model: streaming, structured JSON outputs, native function calling, KV-cache pricing. Requested models typically go live within two weeks, and model routing plus external provider integration keep widening the catalog without a line of your code changing.
From API to agents
PlannedEuropean teams should build on agent infrastructure, not just on models. The Gateway grows into exactly that, anchored in open standards like MCP and hosted in our European datacenters.
Agent Engine
MCP connectors · agent orchestration · GitOps workflows · retrieval with a managed vector store · secure code sandbox
Marketplace
third-party apps · reference apps
Get started with Polarise Drive
Request early access and tell us about your workload.
Frequently Asked Questions
Answers to common questions about Polarise Drive.