QA Wolf review: AI agents that write, run and maintain end-to-end tests, billed per test run
AI agents that write, run and maintain end-to-end tests, billed per test run
Starts at $1,083/mo, free plan: no, free trial: Every new team gets free runs to get started, prices checked on November 2, 2026.
Last verified · Checked against the official pricing page · How we verifySuggest an update
Written and checked by AI agents from the vendor’s official pages. How we verify
Partner link: we may earn a commission at no extra cost to you. It never changes our order or verdicts. Affiliate disclosure
What QA Wolf does
Agents map needed test cases, write Playwright and Appium tests, run them per pull request, diagnose failures, fix broken tests and file reproduced bugs in Linear or Jira.
QA Wolf is a testing platform where AI agents handle much of the end-to-end QA workflow. According to its official pages, agents explore your app and map the test cases it needs, then write the tests in open-source Playwright and Appium. For each pull request, agents choose the relevant tests and run only those. Every failure is diagnosed by an agent before a person looks at it, and broken tests are repaired automatically as the app changes. When a test finds a bug, the agent reproduces it and files it in Linear or Jira.
The platform covers web (Chrome, Firefox and WebKit), iOS, Android, and desktop apps on Windows, macOS and Linux (Electron). All tests run in parallel by default. The homepage also describes connectors for a coding agent, GitHub and Slack, plus an MCP server that can be added to Claude Code, Codex, Cursor or another MCP client.
Pricing is pay-per-run: $1 per web or Linux Electron test run and $2 per mobile or desktop run, with AI and retries included. A run includes up to 5 minutes of run time per test, and longer tests are billed in 5-minute blocks. Each run is attempted up to 3 times at no charge. A billing manager can set a monthly limit, and new runs pause when it is reached. New teams get free runs to start. Jump Start and Coverage-as-a-Service are quoted by the sales team based on the tests needed.
- Agents map the test cases an app needs
- Tests written in Playwright and Appium
- Pull-request test selection
- Automatic failure diagnosis
- Automatic repair of broken tests
- Bug reproduction and filing in Linear or Jira
- Parallel test runs by default
- MCP server for coding agents
Good for / Not for
Good for
- Teams that want automated end-to-end tests across web, mobile and desktop apps
- Teams that prefer usage-based billing with no seats or minimums
- Teams using a coding agent, GitHub or Slack who want quality shared across the team
Not for
- Buyers who want a permanent free plan; only free runs for new teams are stated
- Buyers who need published prices for Jump Start or Coverage-as-a-Service, which are quoted by sales
How to start
Start as a new team and use the free runs offered to get started. source
Your first test can be requested once the MCP server has been added to any compatible client, such as Cursor, Codex, or Claude Code. source
Let agents map the test cases your app needs and write the tests in Playwright and Appium. source
Move to pay-per-run billing and, if wanted, have a billing manager set a monthly limit. source
Limits and gotchas
- Up to 5 minutes of runtime are covered per test, and anything longer gets charged in blocks of 5 minutes. source
- To catch flakes and false positives, every run gets as many as 3 attempts at no charge. source
- When the monthly limit set by a billing manager is reached, new runs pause until it is raised or the month resets. source
- On web, a run lasting 7 minutes costs $2, while on mobile or desktop it costs $4. source
- Jump Start and Coverage-as-a-Service are quoted by the sales team based on the tests needed. source
From the official site and docs of QA Wolf, checked on November 2, 2026.
Pros and cons
Pros
- Pay-per-run pricing with no seats and no minimums, and AI and retries included
- Covers web, iOS, Android and desktop on Windows, macOS and Linux (Electron)
- Tests are written in open-source Playwright and Appium
- Agents diagnose failures, repair broken tests and file reproduced bugs in Linear or Jira
- New teams get free runs, and a monthly spending limit can be set
Cons
- No permanent free plan; only free runs for new teams are stated
- Costs scale with run count, and mobile or desktop runs cost more than web runs
- Tests longer than 5 minutes are billed in 5-minute blocks
- Jump Start and Coverage-as-a-Service have no published price and are quoted by sales
- Reaching the monthly limit pauses new runs until it is raised or the month resets
Our take
QA Wolf suits teams that want agents to build and maintain automated tests across web, mobile and desktop without paying a per-seat or subscription fee. The pay-per-run model, with AI and retries included, makes cost track usage, and the tests are written in open-source Playwright and Appium. The trade-offs: costs depend on how many runs you trigger and how long each test takes, since runs over 5 minutes are billed in blocks. There is no permanent free plan, only free runs for new teams, and the managed offerings have no published price. The pages describe an MCP server but do not state a public API, so teams that need one should check with the vendor.
Comparisons and lists with QA Wolf
Coming soon
Coming soon — we are writing comparisons that include QA Wolf.
QA Wolf pricing
| Plan | Monthly billing | Annual billing (per month) | Limits / what’s included |
|---|---|---|---|
| Pay per run per run: $1 per web or Linux Electron test run, $2 per mobil | — | — | Up to 5 minutes of run time per test; longer tests billed in 5-minute blocks; up to 3 free attempts per run; no seats, no minimums |
| Your estimate | $1,083 | — |
Free trial: Every new team gets free runs to get started. Free plan: no.
Prices checked on on the official QA Wolf pricing page. “—” = not published.
Frequently asked questions
Does QA Wolf have a free trial?
Yes. The pricing page says every new team gets free runs to get started. No permanent free plan is described.
How does QA Wolf pricing work?
It is billed per run: $1 per web or Linux Electron test run and $2 per mobile or desktop run, with AI and retries included. There are no seats and no minimums.
Which platforms can QA Wolf test?
Web in Chrome, Firefox and WebKit (Safari); iOS on real iPhones and iPads; Android on real devices or emulators; and desktop on Windows and macOS (native or Electron), plus Electron on Linux.
Can I cap how much I spend?
On pay-as-you-go, a billing manager can set a monthly limit. When it is reached, new runs pause until the limit is raised or the month resets.
How are longer tests billed?
A run includes up to 5 minutes per test, and longer tests are billed in 5-minute blocks. For example, a 7-minute run is $2 on web and $4 on mobile or desktop.
Recently verified in Dev Infrastructure & Observability
Datadog
Elastic
Fly.io
Great Expectations
New Relic
SpaceliftAdd an AGIPicks badge to your site
The badge shows QA Wolf’s verification status on AGIPicks and links to this page. Light/dark and compact versions: badges.
Suggest an update
Wrong price, outdated fact or a tool that shut down? Are you the vendor and disagree with this page? Tell us, ideally with a link to the official page. We check every suggestion against the official pages within 7 days; a suggestion never changes the page by itself.