Choose the right tier for your needs.
From pay-as-you-go developer inference to fully dedicated sovereign GPU clusters — all on EU infrastructure with zero data retention.
Enterprise
Dedicated and Shared.
Enterprise-grade EU-sovereign GPU infrastructure with predictable pricing. Contact us for a custom quote.
Dedicated Sovereign Inference
Your own isolated GPUs. EU-sovereign. Zero data retention.
- Physically isolated GPU cluster
- Fixed monthly pricing — no usage surprises
- 30+ models with full customisation
- ISO 27001-ready & GDPR Article 44 compliant
- Enterprise SLA with dedicated support
- Dublin, Helsinki, and Paris regions
AI Developers
Pay only for what you use.
Pay only for what you use.
No monthly fees, no minimums.
Prices shown per 1 million tokens. Create an account, buy credits, and start running private inference instantly.
| Description | Context | |||||
|---|---|---|---|---|---|---|
| World's second-strongest open model: 2.4T Alibaba flagship, frontier reasoning, EU-sovereign. | qwen | 256K | $2.50 | $0.63 | $6.00 | |
| The world's strongest open-weight model for coding, agents and cyber. | z-ai | 1M | $1.75 | $0.44 | $4.50 | |
| Moonshot's flagship open model: frontier coding, agents, multimodal, 1M context. | moonshotai | 1M | $3.00 | $0.75 | $15.00 | |
| Fast, low-cost multimodal model with 1M context and frontier reasoning. | z-ai | 1M | $0.20 | $0.05 | $0.50 | |
| Ultra-efficient multimodal model with elite coding from 6B active parameters. | qwen | 256K | $0.20 | $0.05 | $0.50 | |
| 748B multimodal MoE: frontier agentic coding at flash speed. | deepseek | 1M | $0.50 | $0.13 | $1.50 | |
| Efficient 1.6T reasoning model with a lossless 1M-token context window. | deepseek | 1M | $2.00 | $0.50 | $4.00 | |
| Top-3 open model: frontier reasoning at flash speed and price. | deepseek | 1M | $0.25 | $0.06 | $0.30 | |
| Open-weight frontier model with strong coding, agentic performance and a... | z-ai | 1M | $1.50 | $0.38 | $4.50 | |
| Compact 27B open sibling of Qwen3.8 Max: efficient reasoning, EU-sovereign. | qwen | 256K | $0.40 | $0.10 | $2.40 | |
| Open-weight model built for coding, long-context agent work and multimodal... | minimax | 1M | $0.40 | $0.10 | $2.00 | |
| Faster, lower-cost GLM-5 variant tuned for real-time, high-throughput workloads. | z-ai | 198K | $1.20 | $0.30 | $4.00 | |
| Enhanced GLM-5 release with improved reasoning and tool-use performance. | z-ai | 198K | $1.40 | $0.35 | $4.40 | |
| Open-weight model built for end-to-end coding and multi-step agent workflows. | moonshotai | 256K | $1.25 | $0.31 | $4.50 | |
| Multimodal GLM model that handles vision and text for fast,... | z-ai | 198K | $1.20 | $0.30 | $4.00 | |
| Strong general-purpose model from MiniMax, well suited to reasoning and... | minimax | 192K | $0.30 | $0.08 | $1.20 | |
| Compact, cost-effective Qwen model for fast, high-volume general tasks. | qwen | 256K | $0.15 | $0.04 | $0.20 | |
| Large Qwen mixture-of-experts model for advanced reasoning, coding, and multilingual... | qwen | 131K | $0.07 | $0.02 | $0.46 | |
| Qwen embedding model for semantic search, retrieval, and RAG pipelines. | qwen | 40K | $0.02 | - | - |
Prices per 1 million tokens · All inference is zero data retention
Maximize savings
with open models
Select a model, enter your monthly token usage, and see your estimated cost compared to OpenAI GPT-4o. All models run on EU-sovereign infrastructure with zero data retention.
- Drop-in OpenAI compatibility — one line of code
- Zero data retention on every request
- Instant account setup - no sales call required
0% Cheaper
vs. Claude Opus 4.8 at your usage level
≈ 750,000 words
≈ 375,000 words
Cost Comparison
Claude Opus 4.8$0.00
TensorX (-)$0.00
You save$0.00 / month