Weflayr, a French startup founded by former Mangopay CEO Romain Mazeries and Noé Achache, has raised US$2.5 million in Pre-Seed funding. The company is building a platform that helps businesses route every artificial intelligence request to the most appropriate model at the best price. The round was led by 13books Capital, with participation from Founders Future, Kima Ventures, Plug and Play, Station F Ventures, Roxanne Varza, and several angel investors.
A Smarter Approach to AI Model Routing
Hundreds of AI models now compete on quality, speed, and cost, but many companies still rely on a small number of models for most workloads. Weflayr positions itself between enterprise applications and these models as an intelligent routing layer. The platform learns from a client's own production workloads rather than relying on generic benchmarks.
The technology evaluates different models against real data and specific quality and cost objectives. Companies can define routing strategies and approve only the models they trust. Incoming requests are matched to similar previously benchmarked calls so the best-performing approved model handles each task within clearly defined boundaries and cost constraints.
Romain Mazeries noted that payment infrastructure evolved toward sophisticated routing over fifteen years and that AI is now following the same path at much greater speed. AI creators need control and more precision in routing decisions, according to the company's leadership. Weflayr aims to provide that control through an independent intelligence layer for inference.
Early Results and Market Context
Weflayr reports that its first deployment reduced inference costs by more than 55% while maintaining equivalent quality. The company also notably achieved 15 percentage points higher relative quality at equivalent cost compared with a frontier model. These results show how smarter routing can balance performance and economics for AI builders.
The startup enters a market that has already attracted significant attention. OpenRouter, an American platform for AI model routing, was acquired by Stripe last summer in a deal the New York Times valued at US$7.5 billion. Mazeries has compared Weflayr's approach to the routing systems that transformed the payments industry.
Funding and Expansion Plans
The US$2.5 million Pre-Seed round will support engineering and AI research expansion, faster product development, and go-to-market efforts in Europe and the United States. Weflayr is incubated at Station F through the F/AI program dedicated to artificial intelligence. The company is already onboarding its first customers and preparing for broader product availability in the coming weeks.
Investors see AI inference as an emerging infrastructure layer for the next generation of software companies. Mike McFadgen, Partner at 13books Capital, said there is a massive opportunity for an independent intelligence layer that gives AI companies control over quality, performance, and economics. The founders bring experience scaling complex infrastructure from payments to production AI systems.
With fresh capital and a team experienced in critical infrastructure, Weflayr is entering the AI routing market at a time when efficiency and cost control are top priorities for builders. Its focus on evaluating models against real workloads could help companies reduce spending without sacrificing output quality. The startup now plans to scale its platform across Europe and the United States as it works to make AI inference smarter and more controllable.