AI infrastructure your customers can say yes to
Run open models securely on sovereign compute at a fixed monthly price, all behind one OpenAI-compatible API. Built for teams in regulated industries handling sensitive data.
How It Works
One platform for running open models on hardware you trust.
Install a worker on your own GPUs, or use ours. Pull open-weight models onto it. Call them through one API.
Install a worker
One command on Linux, installers for macOS and Windows, or a Docker image. The worker connects out to Pendra. Nothing inbound, nothing exposed. Or skip this step and use our managed UK GPUs, with serving, scaling and failover handled.
Load your models
Pull open-weight models onto your hardware from the console or CLI: Qwen, Llama, DeepSeek, Gemma, gpt-oss and more. Install, swap and pin them without touching the box again.
Call one API
OpenAI- and Anthropic-compatible endpoints with Python and Node.js SDKs. Your existing code just works. Zero retention by architecture, and end-to-end encryption seals prompts to your worker.
Deployment
Run on our GPUs. Or bring your own. Or both.
Pendra is a hybrid platform. Use our managed infrastructure, deploy workers on your own hardware, or mix the two. We orchestrate everything through a single API.
llama3.3:70b
gpt-oss:120b
qwen3.6:35b
deepseek-v4-flash
Developer Experience
Five minutes to first inference.
Drop-in compatible with the OpenAI SDK. Native clients for Python and Node.js. Swap your base URL and your existing code works.
from pendra import Pendra
client = Pendra(api_key="pdr_sk_...")
# Same interface. Sovereign infrastructure.
response = client.chat.completions.create(
model="gpt-oss:120b",
messages=[{
"role": "user",
"content": "Summarise this discharge note."
}]
)
print(response.choices[0].message.content)
Jurisdiction
Your data stays in your jurisdiction. Not even legally reachable.
Sovereignty isn't a region setting — it's the architecture. Run on our UK compute, or on your own GPUs in your own country, under your own law. Either way, no foreign government, including the US, can reach your data.
Use Cases
Your data processing shouldn't stop at the compliance boundary.
Regulated organisations sit on some of the most valuable unstructured data in the world: clinical notes, case files, citizen correspondence, claims documentation. Today, most of it stays locked because the infrastructure to process it privately doesn't exist.
Pendra is the platform for running open models on hardware you trust — ours or yours. Install a model, call it through one API, and keep every token inside your compliance boundary. Teams in healthcare, legal, government and finance build on it where privacy isn't optional.
Clinical document processing
Summarise records, extract from discharge notes, triage correspondence, without data leaving your jurisdiction.
Privileged document review
LLM-assisted review across contracts, briefs and discovery sets on infrastructure that preserves privilege.
Citizen data automation
Classification, redaction and response drafting for FOI requests, benefits processing and case management.
Compliant document intelligence
Process claims, extract KYC data and run agentic workflows under full regulatory control.
Where We're Going
Building the long-term compute layer for private AI.
Pendra isn't just an inference API. We're building the infrastructure stack that makes private, efficient AI processing the default, not the exception.
Now
Next
Future
Managed & hybrid inference
Open-weight models on our UK hardware or your own, delivered as a managed API with end-to-end encrypted private inference.
Advanced security controls
Automatic redaction, granular audit logging, deeper compliance tooling.
Purpose-built inference hardware
Dedicated silicon optimised per architecture. Faster, cheaper inference designed from the chip up.
Get in touch
Building AI products for regulated industries?
If you're building an AI product for healthcare, legal, public sector or finance, we're the inference layer behind it. Sovereign by architecture, compliance-ready, and proven on UK infrastructure, so you can ship without owning the regulatory surface yourself. Let's talk.