Melious
Melious provides an API for open-weight AI models across European infrastructure providers. · Infrastructure
How they use Nebius Token Factory
Melious uses Nebius as one of the European providers serving inference requests through its platform. Its provider documentation lists Nebius in Helsinki. Its subprocessor register and data processing agreement assign Nebius API request processing, open-weight model execution and the return of model responses. Multi-provider routing provides resilience. The register lists prompts, generated answers and technical metadata among the data Nebius processes.
Melious says request bodies and responses remain within the EU and its listed providers do not train on customer traffic. Routing preferences include speed, price and greener power.
Melious offers an API compatible with OpenAI and Anthropic client formats. Its company account also confirms Nebius’s place in its provider network. In a comment on Nebius’s Finland data-center announcement, co-founder Simon Jakobs said Melious planned to use NVIDIA Vera Rubin to deploy open-weight models.
Sources
4 sourcesProviders
18 Oct 2026 · Melious · Web page
Melious lists Nebius in Helsinki among its inference providers. It says request bodies and responses remain within the EU and listed providers do not train on customer traffic. Routing preferences include speed, price and greener power, and responses include provider power usage effectiveness.
Open original sourceSubprocessors — Melious
19 Oct 2026 · www.melious.ai · Web page
Melious lists Nebius as an inference subprocessor processing API requests, running open-weight models and returning responses, with resilience through multi-provider routing. Listed data includes prompts, generated answers and technical metadata such as API keys, timestamps, model parameters and IP addresses.
Open original sourceSimon Jakobs comment on planned Vera Rubin use
18 Oct 2026 · LinkedIn · Web page
On Nebius’s Finland data-center announcement, Melious co-founder Simon Jakobs states planned NVIDIA Vera Rubin use to deploy open-weight models.
Open original sourceAuftragsverarbeitungsvertrag
18 Oct 2026 · Melious · Web page
Melious assigns Nebius API request processing, open-weight model execution and the return of model responses, with resilience through multi-provider routing.
Open original sourceCompany context
melious.aiMelious provides developers and businesses with access to open-weight AI models running across European infrastructure providers. Operated by Melious AI GmbH in Saarbrücken, Germany, its inference API supports text generation, embeddings, reranking, image generation and audio processing through OpenAI- and Anthropic-compatible interfaces. The platform routes requests between providers with automatic failover and offers routing preferences for speed, price and environmental impact. Its products also include tools for AI agents and coding subscriptions for use with existing development tools.
Documented financing · private company
Recent updates
4 updatesMelious adds European inference to Requesty
Melious became available through Requesty's unified API. The integration offers GLM 5.3, GLM 5.3 Flash and DeepSeek models running in European datacenters with zero data retention.
MeliousEcosia discusses using Melious for open-model inference
EU-Startups reported that Ecosia was shifting toward open models accessed through Melious. Ecosia CEO Christian Kroll described future use as probable, while Melious CEO Simon Jakobs said the companies were working together.
EU-StartupsMelious launches European coding subscription
Melious launched a coding subscription offering GLM 5.2 and DeepSeek V4 Flash 0731 on European infrastructure. Developers can connect existing OpenAI- or Anthropic-compatible tools through the Melious API.
Melious