Momentic's launch introduces Mo, an AI agent that bug bashes web and mobile apps on demand for engineering teams that would otherwise run manual QA passes, dogfood sessions, or outsourced test cycles before a release. You point it at a URL or app build with one sentence about what matters, and Mo runs across a hundred or more hosted browsers, simulators, and emulators at once, working through user flows and following each failure until it can reproduce it. Every bug lands with a recording and repro steps , plus logs teams can hand straight to a coding agent to fix.
The launch matters now because scripted end-to-end testing has become a bottleneck for teams shipping AI-generated code faster than humans can maintain the tests around it. CEO Wei-Wei Wu frames the bet plainly: the next frontier of agentic quality is no tests at all, just agents verifying software based on your instructions. In head-to-head evals shared with the launch, Momentic reports Mo finding 3.8x more bugs than Codex paired with Playwright MCP, at 9x the rate per hour and roughly half the cost per bug.
Momentic was founded in 2023 by CEO Wei-Wei Wu, previously an engineer at Assembled and the first engineer at Nashi, and CTO Jeff An, who built cloud and test infrastructure at Splunk, Google, and Robinhood. The company is a Y Combinator W24 alum, and Mo extends its existing agentic quality platform from writing and repairing plain-English tests into fully autonomous exploratory bug hunts across web, iOS, and Android builds.
Comments (15)
Sign in to join the discussion.
P
Priya Katakam14d ago
100 agents bug bashing in parallel is a fun mental image until my staging env catches fire. Do they rate-limit themselves or is it just Lord of the Flies in there?
T
Tomas Lindqvist14d ago
The split-screen showing Mo vs Codex side by side was the right call. Most launch videos would have buried that comparison in a blog post nobody reads.
C
Clara Nwosu14d ago
Impression to engagement on this tweet is doing numbers but the reply-to-quote ratio is suspicious. Are the agents also QAing your engagement?
F
Farrah Osei14d ago
ok wait, 9x more bugs per hour is either a dream or a Jira apocalypse depending on your team size.
D
Dmitri Volkov14d ago
Curious how the eval holds up outside the demo app. Bug counts on your own fixtures always look great, retention on real customer suites is the actual number.
C
Camille Beaufort14d ago
"Catch bugs before they ship" is fine but "Ship like QA never went home" would slap harder on a billboard.
K
Kenji Morisako14d ago
Third AI QA launch this month but the first one where the demo didn't look like a screen recording from 2019. Pacing was tight.
A
Aisha N.14d ago
Every bug comes back with repro steps, logs, AND a video? my human QA doesn't even do that when I bribe them with lunch.
S
sammakes14d ago
The dream of one indie dev shipping without a QA hire feels genuinely closer today. Also mildly threatening if you are a QA hire.
R
Rajiv Menon14d ago
Looks great. One more thing: let me tag flaky tests so Mo triages before filing, otherwise my inbox becomes a swamp.
H
Hannah Weiss14d ago
Before my security team lets this near staging: SOC2? SSO? Can we self-host the agents or are logs shipping to your cloud?
L
Luca Petrangeli14d ago
The Mo wordmark has some rough kerning between the M and o in the hero shot. Small thing but it's the first frame everyone screenshots.
Z
Zayn Mahmood14d ago
This might be the last QA tool designed for humans to configure. Next version, Mo just watches your PRs and files bugs before you open the tab.
S
Sofia Marchetti14d ago
Been watching this team from the sidelines and calling it now, the parallel-agents pattern is going to eat every single-agent QA tool by Q2.
B
Bryan T.14d ago
Naive question: what happens when two agents find the same bug? Do I get 100 duplicate tickets or does Mo dedupe?
Work With Us
Book an intro call to explore our suite of services. We might have a waitlist.
Momentic
Catch bugs before they ship.
Catch bugs before they ship.
Tags
TLVC Rating
Community Rating
No ratings yet
to rate this launch.
Momentic's launch introduces Mo, an AI agent that bug bashes web and mobile apps on demand for engineering teams that would otherwise run manual QA passes, dogfood sessions, or outsourced test cycles before a release. You point it at a URL or app build with one sentence about what matters, and Mo runs across a hundred or more hosted browsers, simulators, and emulators at once, working through user flows and following each failure until it can reproduce it. Every bug lands with a recording and repro steps , plus logs teams can hand straight to a coding agent to fix. The launch matters now because scripted end-to-end testing has become a bottleneck for teams shipping AI-generated code faster than humans can maintain the tests around it. CEO Wei-Wei Wu frames the bet plainly: the next frontier of agentic quality is no tests at all, just agents verifying software based on your instructions. In head-to-head evals shared with the launch, Momentic reports Mo finding 3.8x more bugs than Codex paired with Playwright MCP, at 9x the rate per hour and roughly half the cost per bug. Momentic was founded in 2023 by CEO Wei-Wei Wu, previously an engineer at Assembled and the first engineer at Nashi, and CTO Jeff An, who built cloud and test infrastructure at Splunk, Google, and Robinhood. The company is a Y Combinator W24 alum, and Mo extends its existing agentic quality platform from writing and repairing plain-English tests into fully autonomous exploratory bug hunts across web, iOS, and Android builds.
Comments (15)
100 agents bug bashing in parallel is a fun mental image until my staging env catches fire. Do they rate-limit themselves or is it just Lord of the Flies in there?
The split-screen showing Mo vs Codex side by side was the right call. Most launch videos would have buried that comparison in a blog post nobody reads.
Impression to engagement on this tweet is doing numbers but the reply-to-quote ratio is suspicious. Are the agents also QAing your engagement?
ok wait, 9x more bugs per hour is either a dream or a Jira apocalypse depending on your team size.
Curious how the eval holds up outside the demo app. Bug counts on your own fixtures always look great, retention on real customer suites is the actual number.
"Catch bugs before they ship" is fine but "Ship like QA never went home" would slap harder on a billboard.
Third AI QA launch this month but the first one where the demo didn't look like a screen recording from 2019. Pacing was tight.
Every bug comes back with repro steps, logs, AND a video? my human QA doesn't even do that when I bribe them with lunch.
The dream of one indie dev shipping without a QA hire feels genuinely closer today. Also mildly threatening if you are a QA hire.
Looks great. One more thing: let me tag flaky tests so Mo triages before filing, otherwise my inbox becomes a swamp.
Before my security team lets this near staging: SOC2? SSO? Can we self-host the agents or are logs shipping to your cloud?
The Mo wordmark has some rough kerning between the M and o in the hero shot. Small thing but it's the first frame everyone screenshots.
This might be the last QA tool designed for humans to configure. Next version, Mo just watches your PRs and files bugs before you open the tab.
Been watching this team from the sidelines and calling it now, the parallel-agents pattern is going to eat every single-agent QA tool by Q2.
Naive question: what happens when two agents find the same bug? Do I get 100 duplicate tickets or does Mo dedupe?