LabLab is a community platform for people who build with AI, and its core activity is hackathons. Events are organized together with AI labs and technology vendors, which supply the models, APIs, infrastructure and prize budgets, while LabLab supplies the platform, mentors and streaming operations. Recent and recurring events have been run with sponsors such as AMD, IBM, Vultr, AssemblyAI and Kisaco Research, spanning topics from voice agents and AI trading agents to infrastructure-grade application builds and agentic coding.
The structure of an event is consistent from the platform's side: an opening kick-off stream presents the technology and schedule, teams then build during a defined window, mentors are reachable on demand through Discord, and the event closes with streamed pitches where each team presents and takes questions from mentors and judges. Formats range from 48-hour sprints to month-long challenges, and events can be fully online, hybrid or on-site, for example an invited on-site phase at a venue like the Santa Clara Convention Center or Pierpont Place in Salt Lake City.
The platform deliberately avoids prescribing a single problem to solve. Instead it asks participants to innovate on top of a given sponsor technology, with the expectation that teams come from mixed backgrounds: engineers, designers, data scientists and domain experts. Beyond events, LabLab publishes AI tutorials and articles, maintains a prototype gallery and leaderboard, and runs community channels on Discord, Twitch, YouTube and LinkedIn. Registration and attendance for hackathons are free; sponsor programs tied to specific events, such as the AMD AI Developer Program, may require their own account.