Pricing.
Predictable pricing for secure access to AI infrastructure. Contact for custom pricing.
Model pricing
| Model | Parameters | Input / 1M tokens (€) | Output / 1M tokens (€) | Cache read / 1M tokens (€) |
|---|---|---|---|---|
| GLM-5.3-Flash | 320B · 18B active | €0.30 | €0.90 | €0.035 |
| DeepSeek-V4-Flash | 284B · 13B active | €0.25 | €0.75 | €0.035 |
| Qwen3.8-27B | 27B (dense) | €0.50 | €3.00 | €0.05 |
| Qwen3-Embedding-4B | 4B (dense) | €0.05 / 1M tokens | ||
Secure sandboxes, billed by the hour.
Isolated, firewalled environments where your agents execute code and tools — next to your inference. Every sandbox ships with a comprehensive list of preinstalled packages for document processing, data analysis, and engineering.
| Specification | Rate |
|---|---|
| vCPU | €0.02 / vCPU-hour |
| RAM | €0.02 / GB-hour |
| Temporary 10GB storage | Included |
| Network | 1 Gbps included |
Sandbox 1
1 vCPU / 1GB RAM
€0.04 /hour
Lightweight document processing and everyday agent tasks.
Sandbox 2
1 vCPU / 2GB RAM
€0.06 /hour
Data analysis with the full preinstalled package list.
Sandbox 3
2 vCPU / 4GB RAM
€0.12 /hour
Engineering-grade compute for heavy workloads.
Retrieval-augmented generation
Built-in retrieval pipeline for grounding model responses in your documents — embeddings, semantic search, and reranking, all inside the EU.
| RAG | Price |
|---|---|
| Retrieval query | €0.002 / query |
| Embedding (AI) | €0.05 / 1M tokens |
| Rerank (AI) | €0.03 / 1M tokens |
| Storage | €0.20 / GB/month |