Teamwork: When AI Becomes a Research Partner | Google Antigravity Blog
The orchestration framework coordinates multiple Gemini agents over hours or days across five specialized patterns and is available on all paid plans.
- Ships specialized patterns for Iterative Coding, Distributed Coding, Long Proof, Self-Verification, and Document Review.
- Solved seven open research problems in mathematics and theoretical computer science, including Knuth's Cycles Conjecture verified in Lean.
- Built a cycle-accurate RISC-V CPU simulator from scratch that boots xv6 with a 0.71% cycle alignment error against BOOM.
- Generated code optimizations merged directly upstream into open-source libraries Eigen and ParlayHash.
- Reached 71% on Google's internal TCSBench evaluation using Gemini 3.7 Flash with 3.1 Pro.
Paid Antigravity users can run automated multi-agent teams that critique and iterate across long-horizon research, systems engineering, and coding tasks.
