Sign Up

Docs

Wire up the stack.

Five SDKs, one deadline. Quickstarts and official docs for every sponsor — Composio, Nebius Token Factory, Nebius AI Cloud, Tavily, OpenClaw — plus the BuilderShip platform docs.

Managed inference for open-weight models

Nebius Token Factory

OpenAI-compatible API in front of dozens of open-weight models — Llama, Qwen, DeepSeek, Mistral, and more. Drop your existing OpenAI client at the Nebius base URL, change the API key, ship.

  • OpenAI-compatible /v1/chat/completions and /v1/embeddings
  • Streaming, tool use, and JSON mode supported
  • Per-builder credits — keys auto-rotate post-event
  • Base URL: https://api.studio.nebius.ai/v1

GPU compute & serverless deploy

Nebius AI Cloud

Bare-metal H100/H200/A100 instances when you need raw training power, plus Nebius Serverless for one-command deploys of agents and APIs. Builders get GPU credits scoped to the event.

  • GPU instances from H100 to A100 — pay by the minute
  • Nebius Serverless for stateless agent deploys
  • Object storage + managed Postgres in the same region
  • Console at console.nebius.com

Search & web extraction for agents

Tavily

Real-time search API tuned for LLMs — clean, ranked, citable results plus structured web extraction. The retrieval layer for any agent that needs to know what just happened on the internet.

  • /search returns ranked passages with citations
  • /extract pulls clean structured content from a URL
  • Sub-second latency, designed for agent loops
  • Free credits across the build period

Open agent runtime

OpenClaw

Open-source agent framework for building, deploying, and operating agents on your terms. Local-first install, ships to Nebius Serverless in one command, plays nicely with Composio and Tavily out of the box.

  • Local install — one binary, no cloud dependency
  • Skill packs for Claude Code workflows
  • Sandboxed tool execution by default
  • Deploys to Nebius Serverless with one command