Choose a starting point for your work
Four products. Pick the one that fits how you want to build.AI Cloud
Spin up GPU VMs and multi-node clusters for training and custom stacks, with full control.Token Factory
Serve open-source models via an OpenAI-compatible API with real-time and batch inference, dedicated endpoints, and production SLAs.Serverless
Build and deploy serverless AI jobs and endpoints from your container in minutes — no Kubernetes, no infrastructure.Tavily
Real-time web search and content extraction for LLMs and agents via API and SDKs.Bring Nebius into the editor you already use.
Drop-in skills and MCP servers for Claude Code, Cursor, and OpenAI clients. Same Nebius credentials; your agent just gets more useful.Claude Code
Drop in the open-source Nebius Skill so Claude Code knows Token Factory, AI Cloud, and Serverless. Add the Tavily MCP server to search the live web from your shell.Cursor
Wire Token Factory in as a custom model provider, and add the Tavily MCP server for in-editor web search and extraction.Tavily Agent Skills
Pre-built skills that give your agent web search, extraction, and crawling out of the box.View docs →OpenAI Codex
Configure OpenAI Codex to use Nebius coding models, and add the Tavily MCP server for live web search inside your terminal.VS Code (Github Copilot)
Hugging Face's VS Code Chat extension — routes Github Copilot through Nebius Token Factory models.View extension →Cline
Open-source AI coding agent for VSCode + JetBrains. Direct access to Nebius coding models.View docs →Continue
Open-source autopilot for VS Code & JetBrains, pointed at Nebius.View docs →Kilo Code
Multi-mode coding agent for VS Code on Token Factory models.View docs →Zed
Configure Zed's inline assistant against Token Factory models.View docs →What can you build
A few examples to get you going. More in the docs and cookbooks.Inference
Serve open models in production with a simple API, real-time or batch inference, dedicated endpoints, and autoscaling for traffic spikes.View guides ↑Training & Fine-tuning
Train new models or fine-tune existing ones with scalable GPU jobs, track experiments, and push the best checkpoint into production.View guides ↑Industries AI Applications
Build domain-specific workflows in healthcare, robotics, and more with secure data access, guardrails, and reusable patterns for RAG, copilots, and automation.View guides ↑Join the community
Get help, share what you’re building, and connect with other Nebius builders.Discord
Ask questions, share what you're building, and get help from other Nebius builders and the team.Join Discord ↑Hackathons & Events
Hands-on sessions for inference, training, and real workloads. Bring a laptop, ship something real.See upcoming events →GitHub
Examples, templates, and reference architectures to copy-paste into your stack.Go to repos ↑YouTube
Walkthroughs, demos, and technical deep dives from the Nebius team.See videos ↑X / Twitter
Daily updates, release threads, and the occasional GPU benchmark from @nebiusai.Follow @nebiusai ↑Three ways to get more from Nebius
Credits, infrastructure support, and standing in the community.Builders Network
Run events, post tutorials, build with Nebius. Earn up to $2,600 of credits as you climb tiers. Get promoted to Contributor and Ambassador on points.Learn more →Startup Program
Credits, guidance, and infrastructure support for teams building on Nebius.More ↑Nebius Academy
Expert-led courses for developers, engineers, and teams — agentic development, AI performance engineering, onboarding, and more. Free and hybrid formats.More ↑Pick the tools you already love. They work on Nebius.
First-party integration docs for the frameworks, gateways, and orchestrators we've tested end-to-end. A few standouts below — the rest are one click away.Get in touch
Run your next event with us.
Tell us what you’re planning. We’ll come back inside 24 hours with a venue, a partner shortlist, and a budget.- San Francisco · NYC · London · Remote
- builders@nebius.com
- Accepting Q3 bookings
Build in public