Momentic launches bug-catching AI agents
The gist
San Francisco testing startup Momentic launched Mo, an AI QA agent, alongside a public bug-finding benchmark.
The launch drew a mostly skeptical Hacker News thread over cost and defensibility.
How it unfolded
- Nov 2, 2026Momentic launches on Hacker News with the post "Launch HN: Momentic (YC W24) – Agents that catch bugs by using your app."
- Nov 2, 2026Alongside the launch, Momentic releases QA-Bench, a public benchmark for how well AI agents find bugs in real apps, with method and data published on Hugging Face.
- Nov 2, 2026Wei-Wei Wu says Mo found 2.7x more reproducible bugs than Codex and 6.3x more than Claude Code, using the same apps, time budget and prompt for every agent.