RelaxAI – UK sovereign LLM inference at 80% cheaper than OpenAI/Claude
RelaxAI introduces a UK-based LLM inference service, promising 80% cost savings over OpenAI/Claude and compliance for data residency. While its OpenAI-compatible API and use of open-source models attract interest, the community is highly skeptical of the aggressive cost claims and the vague marketing around "UK sovereign" data. Discussion revolves around the true value proposition against established providers and the implications of data location.
The Lowdown
RelaxAI has launched a new API offering UK sovereign LLM inference, aiming to provide a cost-effective and data-compliant alternative for developers. The service claims to be up to 80% cheaper than major providers like OpenAI and Claude.
- The API provides official SDKs in Python, TypeScript, and Go, with 1:1 compatibility with the OpenAI API for seamless integration.
- It leverages popular open-source models, including Kimi K2.6, DeepSeek V4 Pro, Nemotron 3 Super, and GPT OSS 120b.
- The inference is run on NVIDIA Blackwell GPUs housed within UK sovereign cloud infrastructure, ensuring data never leaves UK jurisdiction.
- This setup is particularly targeted at regulated sectors such as finance, legal, health, and defense, which have strict data residency requirements.
- Despite the regulatory benefits, the company notes that many early users are primarily drawn to the significant cost savings.
RelaxAI positions itself as a vital option for organizations needing secure, localized, and affordable access to advanced natural language processing capabilities.
The Gossip
Sovereignty Scrutiny
Commenters critically examined the term "UK sovereign," questioning its true meaning and value. Skepticism arose regarding data privacy in the UK, citing the Online Safety Act and the country's involvement in intelligence-sharing alliances like Five Eyes. The discussion delved into whether "sovereign" was a genuine advantage or merely marketing jargon, with some suggesting it implied government control.
Pricing & Performance Ponderings
The claim of being "80% cheaper" sparked significant debate. Many questioned the fairness of the comparison, suggesting that the cost reduction might simply reflect the use of smaller or less powerful open-source models compared to OpenAI's frontier models. The community also compared RelaxAI's offering to services like OpenRouter, which aggregates various models, and questioned the economic viability of running mid-tier models on UK hardware.
Marketing & Messaging Misfires
The company's branding and communication strategy received criticism. Users found the name "RelaxAI" to be uninspired and the marketing tagline "a whole new way of thinking" to be generic fluff. There was also feedback that linking directly to API documentation rather than a main website hindered immediate context and understanding for new visitors.
Foundational Facts & Figures
Many commenters requested more transparent and detailed information about RelaxAI. Specific questions arose regarding the identity and background of Civo (the parent company), the precise location of their data centers, corporate structure, and clearer breakdowns of pricing for different use cases. The absence of specific model details on the main site was also noted.