The news
Momentic, a San Francisco software testing startup, unveiled Mo on September 28, 2026, an AI testing agent that checks web and mobile apps without developers writing or maintaining test scripts, according to SiliconANGLE. Early users include Notion.
Developers give Mo an app's web address, a description of what to test and test login credentials, using a chat interface or a command-line tool. Mo then spins up a group of AI agents that try thousands of variations and edge cases, SiliconANGLE reported. It returns a report on what was tested and what failed, with reproduction steps and a video of each failure. It can also read requirements from Jira tickets, Linear issues or Confluence documents.
Momentic's product documentation describes separate agents that plan coverage, explore the product, hunt for bugs and reproduce them. Each suspected bug is rechecked from a clean starting state before it is reported, a step meant to filter out false alarms. Mo works on web apps in a Chromium browser and on iOS simulators and Android emulators, but not on physical devices, according to the documentation.
Wei-Wei Wu, Momentic's co-founder and CEO, told SiliconANGLE that any script kept in a codebase is something a person has to maintain. "Our bet is that there's no tests in the future," Wu said. Teams that still want a conventional test suite can convert Mo's findings into the script format of their choice, according to Wu.
Mo had been in customer testing for the past few weeks with companies including Notion, Superpower, Iris, Committee for Children and Boundless, SiliconANGLE reported. Momentic has not published pricing for Mo. Its website lists a free tier and pay-as-you-go plans for its broader platform starting at $125 a month plus usage.
Wu and chief technology officer Jeff An founded Momentic in 2023, and the company went through Y Combinator's Winter 2024 batch, according to its website. In June 2026, it launched what it called an agentic quality platform, with agents that expand test coverage and sort test failures. That release listed investors including Standard Capital, Y Combinator and Dropbox Ventures.
The numbers
- Launch date
- September 28, 2026
- Named early Mo customers (per SiliconANGLE)
- 5
- Platform pay-as-you-go plan (company website)
- From $125 a month plus usage
- Bugs caught before production (company website counter, September 29, 2026)
- 117,010
- Founded
- 2023
Why CEOs should care
For CTOs and engineering leaders, the pitch targets a bottleneck SiliconANGLE described: AI tools are speeding up how fast code gets written, and that code still needs checking before release. Before retiring any scripts, run Mo alongside your existing suite on a release candidate. Ask Momentic for its false-positive rate, how Mo handles login flows and test data, and how its reports plug into your continuous integration pipeline, the automated system that checks code before release.
CFOs should expect the cost model for quality assurance to change. Instead of paying people to write and repair scripts, companies would pay for agent runs, which scale with usage. Momentic has not disclosed Mo pricing, so ask for per-session costs, spending caps and a forecast based on your release schedule. Compare that with the staff time and tool licenses your current test suite consumes.
CISOs and compliance teams have two issues to settle. Mo takes test credentials, and Momentic's website says sessions can run with an organization's MCP servers and secrets attached; MCP is a standard for connecting AI agents to tools and data. Limit Mo to dedicated test accounts with the fewest permissions possible and keep production data out. In regulated industries, confirm that session reports and video recordings give auditors enough evidence of what was tested, and check how long Momentic retains them.
The bigger picture
Scriptless testing trades one kind of risk for another. A script runs the same steps every time, which makes results easy to repeat and audit. An agent that explores an app can find problems no one thought to script, but its path may differ between runs. Momentic's answers are a separate agent that confirms each bug from a clean state and the option to export findings as conventional scripts. How well that balance holds up in large, fast-changing codebases will decide whether buyers treat agentic testing as a replacement for scripted suites or as an extra layer on top of them.
What’s next
SiliconANGLE did not report a general availability date or pricing for Mo, and the documentation notes it does not yet support physical devices. Watch for published pricing, results from the early customers, and whether Momentic adds real-device testing, which many mobile teams rely on before release.
What “Fact-checked” means
Fact-checking means testing a story’s facts against the evidence before it is published. This story went through at least two separate checks before this version was published.
- What we checked
- Its names, figures, dates, job titles, quotes and who said what were checked against the story’s sources, including its main source where it could be opened. The headline was checked for accuracy and overstatement.
- How
- A first check reviewed the whole story. If it passed, a second, skeptical check went back to the sources to look for mistakes in the most important facts. If a check flagged the story, it was edited to fix the problems found, and a separate re-check then reviewed the whole story again.
- Who
- The checks are made by our newsroom, as steps kept separate from the writing, under rules set by our editor, Hussein Mukhtar. A story the checks still flag is held for the editor, who decides whether it is fixed, published or dropped.
- If something is wrong
- “Fact-checked” does not mean error-free. If a material error is found after publication, we correct the story and add a note saying what changed. Report an error
