Swiss AI lab Giotto has launched a new family of reasoning models — Giotto Alto and Giotto Solo — built around what it calls sovereign AI: advanced models that run on European cloud infrastructure or on the user's own hardware, with no dependency on US cloud services. The announcement, made in Lausanne on October 27, 2026, positions the flagship model as free for everyone.
Two models, three deployment options
Giotto Alto is the family's flagship, aimed at demanding reasoning and enterprise workloads. It runs on eight B300 GPUs and is offered free to everyone through Giotto's shared cloud. Giotto Solo is the smaller model, designed to run on a single GPU; the company describes it as the smartest model that runs on one GPU, targeting organisations and individuals that want to keep their AI on infrastructure they control.
Across the family, Giotto offers three deployment options:
- Shared cloud, hosted and operated in Europe.
- Dedicated cloud in isolated environments at partner data centers in Switzerland or elsewhere in Europe.
- On-premises installation on compatible hardware through partner integrators, ranging from a one-GPU workstation for Solo to multi-GPU server racks for Alto.
On pricing, shared cloud access is free. API access on the shared cloud uses a flat fee for all-you-can-use requests, limited to one concurrent request at a time, with token packages available for multi-concurrency — a departure from uncapped per-token pricing. Dedicated cloud instances, starting at one GPU, are rented monthly and include the Giotto software licence. On-premises deployments are licensed annually or over multiple years, with pricing based on the number and type of GPUs.
Giotto is also launching its app for iOS and Android alongside a free anonymous chat that requires no account. The company states its cloud services are GDPR-aligned and do not depend on US cloud infrastructure.
Benchmark numbers, self-reported
Giotto publishes benchmark comparisons on its own website. The figures below are the company's self-reported results, not independently verified third-party scores:
- Giotto Alto: 100.0% on AIME 2026, 99.80% on Math-500, 91.9% on GPQA-Diamond, and 52.0% on HLE.
- Giotto Solo: 100.0% on AIME 2025, 96.6% on AIME 2026, 99.80% on Math-500, 88.3% on GPQA-Diamond, and 37.5% on HLE.
In the same tables, Giotto compares its models against specific variants from OpenAI, Anthropic, Google, Moonshot AI, Mistral, and others. Readers should treat the "smartest European sovereign AI" framing as the company's own positioning claim rather than a verified ranking.
Context
Giotto was founded in Lausanne in 2017 by EPFL mathematics alumni and has worked on machine-learning interpretability, language models, and reasoning systems. In the ARC-AGI benchmark competitions, its results placed 11th in 2024 and second in 2025, among roughly 1,500 teams.
CEO and co-founder Aldo Podestà framed the launch around European control of AI infrastructure: "Europe must offer more than a different flag on the same dependency. Sovereignty has to belong to the people and organisations using AI, and not just to the companies building it." The launch lands in a period of intense competition among European AI efforts, and its combination of a free flagship, flat-fee API pricing, and single-GPU deployment directly targets both individual users and organisations that want to host advanced AI in-house.
Sources
- Launch announcement (Solo + Alto, October 27, 2026, Lausanne); Alto on eight B300 GPUs, free; three deployment options; pricing structure; iOS/Android app, GDPR alignment; founding and ARC-AGI results; CEO statement: Giotto.ai SA press release via EQS News
- Benchmark figures (AIME, Math-500, GPQA-Diamond, HLE) — self-reported by Giotto: giotto.ai