Scaleway launches Generative APIs, a drop-in alternative to OpenAI APIs, 100% made in Europe

Scaleway, the European cloud provider, today announces the public launch of Scaleway Generative APIs, its new service designed to democratize access to generative AI while ensuring European technological sovereignty. This innovative offering provides enterprises, startups, and public administrations with state-of-the-art AI capabilities while guaranteeing that data never leaves Europe.

European values at the core of AI Innovation

Unlike hyperscaler solutions, Scaleway Generative APIs ensure complete data sovereignty within European borders, adhering to the highest standards of continental data protection laws, including GDPR. The service is built entirely on European infrastructure, in particular hosting AI models in low-carbon data centers in Paris.

At the core of the service lie the following values:

• Empowering full data privacy: Scaleway AI services were built to not collect, read, reuse, or analyze the content generated by the APIs
• Fostering an open, collaborative ecosystem: Scaleway is distributing open science and open-weight models from leading research labs, wherever they come from
• Refusing vendor lock-in: Scaleway is making every possible effort to reach interoperability with the AI-native open source software developers love.

Seamless integration, no trade-off on performance

Designed as a drop-in replacement for existing AI-native applications, Scaleway Generative APIs offers:

• Complete compatibility with OpenAI SDKs and Langchain
• Sub-200ms latency for European users
• Built-in JSON schema for structured outputs
• Native function calling capabilities
• Comprehensive RAG (Retrieval-Augmented Generation) stack
• Advanced analytics dashboard

European companies using OpenAI's APIs (via its platform or its partner Azure) can switch to Scaleway with minimal code changes. It’s this simple:

Real results from real users

Since early 2024, Scaleway has worked closely with select partners across various sectors to refine and validate the service. Early adoption has demonstrated impressive results.

One AI startup user reported on the platform's scalability:
“The results were impressive: in 3 hours and 18 minutes, the latest Meta Llama model processed around 50,000 files. The processing speed was remarkable.”

Another tester highlighted how he was able to significantly simplify its architecture:
“Using the latest vision model by Mistral for analyzing images and automatically generating structured outputs saved us from using 3 different AI models for the same task.”

Pricing and availability

The service is now available in public beta, free of charge to all Scaleway customers without limitations. Post-beta pricing will start in Q1 2025, following a competitive pay-as-you-go model based on token usage.

Organizations can begin using Scaleway Generative APIs immediately by:

• Accessing the Scaleway Console
• Obtaining API credentials
• Exploring the model catalog
• Integrating with existing applications

For more information about Scaleway Generative APIs, visit scaleway.com/en/generative-apis or join the Scaleway community on Slack (#ai channel) to connect with other developers and builders.