How much does Momentic cost?

Free includes $50 of usage a month, five concurrent browsers, two Android and two iOS devices and 30 day result retention. Scale is $500 a month billed in advance as $500 of usage, with top ups of $50 to $1,000 up to $1,000 a month that never expire, 25 browsers and ten devices. Enterprise is custom, priced per test on an annual invoiced contract with SAML, SCIM, audit logs and an SLA. Usage is metered per step, browser minute, emulator minute and AI call, and runs pause at a zero balance.

What are the best alternatives to Momentic?

Based on documented capability coverage across the Agentic Index 14 point taxonomy, the closest alternatives to Momentic among agent infrastructure platforms are OpenAI Agents platform, Cloudflare, Mastra, Agno. Each is graded on the same public evidence, and no vendor pays for placement.

Does Momentic support MCP and tool calling?

Full under the Agentic Index verification standard. Momentic publishes public evidence of Model Context Protocol (MCP) support, tool calling, and API or SDK extensibility for connecting agents to external tools. Evidence basis: A hosted MCP server signs in with OAuth 2.1 and offers tools to start, message, stop and list Mo sessions, upload builds, and list, triage and retest findings, and a local MCP server running on an API key adds 24 tools that create and edit. Graded from public evidence only against the Agentic Index 14 point capability taxonomy. No vendor pays for placement.

Does Momentic have SOC 2, SSO and role based access control?

Full under the Agentic Index verification standard. Momentic publishes public evidence of enterprise security and identity controls, covering a compliance posture such as SOC 2 or ISO 27001, single sign on (SSO or SAML), and role based access control (RBAC). Evidence basis: Enterprise adds SAML and OIDC single sign on, SCIM provisioning and audit logs of account and test activity, while Google sign in comes with every plan and accounts carry Admin and Member roles. Model providers work under zero data retentio. Graded from public evidence only against the Agentic Index 14 point capability taxonomy. No vendor pays for placement.

Does Momentic give you tracing and observability into agent runs?

Full under the Agentic Index verification standard. Momentic publishes public evidence of tracing, observability and audit logging into agent runs, so a team can see what an agent did and why. Evidence basis: Each run keeps video, step traces, screenshots, console and network logs and the history of any auto heal, and results export as JUnit, JSON, Allure or Playwright reports, with a result viewer the customer can host. Mo shows a transcript fo. Graded from public evidence only against the Agentic Index 14 point capability taxonomy. No vendor pays for placement.

Back to vendors

Momentic

Also known as: Momentic.ai, Momentic, Inc., Momentic AI, Momentic Mo

Visit site

Agent infrastructureindependentVerified 2026-11-08

Entry priceFree with $50 of usage a month, or Scale at $500 a month billed in advance as $500 of usage, with usage metered per step, minute and AI call.Full pricing detail

Momentic is an AI QA platform from Momentic, Inc. of San Francisco whose agent, Mo, explores web, iOS and Android apps in hosted browsers and emulators, files reproduced bugs, reviews pull requests by testing preview builds, and writes regression tests as draft pull requests. Repository tests in YAML run in CI with self healing that commits repairs to the branch. Built to check the changes coding agents ship, it has a free plan with $50 of monthly usage and a Scale plan at $500 a month.

Momentic, Inc. is a San Francisco company founded in 2023 by Wei-Wei Wu and Jeff An and backed by Y Combinator, FCVC, General Catalyst, Standard Capital and Dropbox Ventures. It sells AI testing for teams whose code is increasingly written by agents, through two products on one platform: Mo, an AI QA agent that tests running apps, and Momentic repository tests, plain YAML files that live in the customer's repository and run through the momentic and momentic-mobile command line tools. Notion, Xero, Webflow, Whatnot, Retool and Quora are among the customers on its site.

A Mo session starts from the web app, the qa command line tool, an API call, a schedule, an @momentic-ai comment or a label on a pull request, or an MCP client such as Claude Code or ChatGPT. A session agent sends explore, bug bash and reproducer agents into their own hosted Chromium browsers, iOS simulators or Android emulators, and at higher effort levels coordinating agents split the app by area, with a reproducer confirming each functional bug before it is reported.

Findings land in a review queue for a person to accept or dismiss, accepted bugs can file Linear issues on their own, and on pull requests Mo posts a review and a pass or fail check. When a repository's regression setting is on, Mo's Coder writes a Momentic test for a bug and opens it as a draft pull request; it writes tests, not application fixes. A knowledge base of terms, rules, flows and code insights, retrieved by embeddings, and a memory of past locators and assertions that expires after 30 idle days shape later runs, while auto heal triages failing repository tests and delivers verified repairs.

The Free plan includes $50 of usage a month with five concurrent browsers and four mobile devices, Scale costs $500 a month as prepaid usage with top ups of up to $1,000, and Enterprise is priced per test on an annual contract. Usage is metered at $0.0125 a test step, $0.025 an AI step, $0.005 a browser minute and from $0.10 an emulator minute, with triage at $1.25 a call, and runs pause when the balance reaches zero.

Google sign in comes with every plan, while SAML and OIDC single sign on, SCIM and audit logs come with Enterprise; the security page calls SOC 2 Type 2 certified while the trust center lists it in progress. Hosted browsers run in two US regions, emulators can be pinned to the US, Europe or Asia, repository tests can run locally or in the customer's CI, and tunnels with fixed egress IPs reach private staging sites, while the control plane stays in Momentic's cloud.

Momentic fits product and engineering teams shipping web or mobile apps quickly, often with coding agents, that want an agent to explore releases, catch regressions on every pull request and keep test suites healing without a QA team writing scripts. It fits less well for teams that need Firefox, Safari or physical devices, an on premises edition, a fixed monthly bill, or an agent that repairs application code rather than tests.

Vendor details

Canonical URL

https://momentic.ai

Subcategory

AI QA agent that explores web and mobile apps, files reproduced bugs, reviews pull requests and maintains self healing YAML tests in CI

Funding status

Venture backed; a $3.7M seed led by FCVC and a $15M Series A led by Standard Capital, with Y Combinator, General Catalyst and Dropbox Ventures among its investors.

Company status

independent

Use cases & customers

Primary use cases

exploratory bug bashes on a URL or mobile buildQA review and checks on every pull requestself healing end to end tests in CIturning reproduced bugs into regression teststriage of failing tests into bugs, flakes and environment issues

Target customers

product engineering teams shipping web apps with coding agentsmobile teams testing iOS and Android buildsstartups without a dedicated QA teamenterprises that need SSO, SCIM and audit logs for testing

Deployment options

Momentic cloud with hosted Chromium browsers in two US regionsHosted iOS simulators and Android emulators pinned to the US, Europe or AsiaRepository tests on local browsers and emulators or in the customer's CIPrivate network access through the Momentic connector, tunnels and fixed egress IPsSelf hosted result viewer

Integrations

GitHub App for pull request reviews, checks, comments, healing pull requests and commits; GitLab read access; Linear issue filing; Jira and Confluence through Atlassian Rovo MCP; Slack notifications; Notion, Vercel and Cloudflare; custom MCP servers with OAuth or header auth; CI on GitHub Actions, GitLab CI, CircleCI, Buildkite, Jenkins, Azure DevOps and Travis; email inboxes and SMS numbers for tests; a REST API and hosted and local MCP servers.

In practice

A startup installs the Momentic GitHub App and labels a pull request written by its coding agent; Mo tests the preview deployment, reproduces a broken checkout step, posts a failing check with a video, and opens a draft pull request with a regression test once the team accepts the bug.

A mobile team uploads a nightly Android build and schedules a Mo bug bash at high effort, then accepts the reproduced crashes in the review queue so they file as Linear issues automatically.

A platform team runs YAML tests in GitHub Actions on the Scale plan; when a renamed button breaks three tests, auto heal classifies the failure, verifies a repair in a live browser and commits the fix to the branch.

Sources & related URLs

Related / legacy domains

https://momentic.ai/docs https://momentic.ai/mo https://momentic.ai/pricing https://momentic.ai/security https://trust.momentic.ai

Research sources

Agentic Index coverage score

12.5 / 14 capabilities · 89%

Integrations & Tool Calling	Full
The GitHub App posts pull request reviews, comments and a pass or fail check, opens healing pull requests and commits test repairs to branches, and Mo files Linear issues at the press of a button or on its own once a bug is accepted. Results from repository tests can raise Jira issues, notices go to Slack, and Mo reads Jira, Confluence, Notion, Vercel and Cloudflare. Teams can attach any MCP server with OAuth or header auth, and Mo can use its write tools. GitLab access is read only. Sourcemomentic.ai/docs/mo/integrations/mcp.mdread 2026-11-08
Workflow Orchestration	Full
Each Mo session runs a session agent that dispatches explore, bug bash and reproducer agents, each in its own browser or device, and at high and max effort coordinating agents divide the app by area. Effort levels set a budget from about 15 minutes to two hours, and setup and teardown endpoints named in the brief run around the session. Repository tests run across parallel workers and shards, reuse shared modules, and can include goal based AI steps that plan their own actions. Sourcemomentic.ai/docs/mo/sessions.mdread 2026-11-08
Knowledge Grounding & RAG	Full
A knowledge base holds terminology, agent rules, user flows and code insight entries the platform generates, scoped to the organization, an app, an environment or a single test and retrieved by embeddings while the agents work. Mo also reads selected repositories as read only checkouts with about two weeks of history, and AI test selection builds a local index of the code to choose which tests a change affects. Sourcemomentic.ai/docs/ai/knowledge-base.mdread 2026-11-08
Human Oversight & Guardrails	Full
Every Mo finding starts in a review queue where a person accepts or dismisses it, and a pull request can be allowed through despite a finding. Regression tests arrive as draft pull requests, auto heal delivers repairs as a pull request by default with direct commits as an opt in mode and reviewers or assignees set per repository, and knowledge base suggestions wait for approval. Mo asks clarifying questions in the session chat, and inside Gemini it confirms before any write action. Sourcemomentic.ai/docs/mo/triage.mdread 2026-11-08
Security, Identity & Governance	Full
Enterprise adds SAML and OIDC single sign on, SCIM provisioning and audit logs of account and test activity, while Google sign in comes with every plan and accounts carry Admin and Member roles. Model providers work under zero data retention agreements, and Enterprise customers can opt out of having their data used to improve Momentic's agents. The security page lists SOC 2 Type 2 as certified, while the Oneleet trust center shows it in progress beside an earlier audit report and a penetration test, all on request. Sourcemomentic.ai/securityread 2026-11-08
Observability & Auditability	Full
Each run keeps video, step traces, screenshots, console and network logs and the history of any auto heal, and results export as JUnit, JSON, Allure or Playwright reports, with a result viewer the customer can host. Mo shows a transcript for every sub agent, and its report bundle carries bugs, verdicts, triage decisions, video and session replays. Enterprise adds audit logs of account and test changes, and analytics keep pass rates and flake history. Sourcemomentic.ai/docs/mo/reports.mdread 2026-11-08
Memory & State Persistence	Full
A memory feature stores past locator and assertion traces for each organization so later runs resolve the same elements faster, expires entries after 30 days without use, and switches on or off per test or across a project. Marking a finding as working as intended drafts a knowledge base entry that a person accepts, and entries are versioned, can be disabled or archived, and shape later sessions. Deleting account data goes through support. Sourcemomentic.ai/docs/ai/memory.mdread 2026-11-08
Deployment & Data Residency	Partial
Hosted browsers run in two US regions, hosted iOS simulators and Android emulators can be pinned to the US, Europe or Asia, and customer data is kept in the US. Repository tests also run on local browsers and emulators or inside the customer's own CI, and a connector, tunnels and fixed egress IPs reach private staging sites. The control plane and every Mo session stay in Momentic's cloud on Google Cloud, with no on premises edition. Sourcemomentic.ai/docs/running-tests/hosted-test-environments.mdread 2026-11-08
Prebuilt Agents / Templates / Packs	Partial
Teams build their own reusable modules, groups of steps shared across tests, and an MCP tool suggests existing modules when an agent writes a test. Five agent skills for coding agents install from a published index, and CI templates cover the common providers. Mo's explore, bug bash and reproducer agents come inside every session rather than as agents a team selects, and no catalog of prebuilt tests or flows is offered to adopt. Sourcemomentic.ai/docs/mo/sessions.mdread 2026-11-08
Triggers & Channel Coverage	Full
A Mo session starts from the web app, the qa command line tool, an HTTP call that opens a session or a pull request session, a cron schedule, an @momentic-ai comment or a label on a pull request, or the hosted MCP server from Claude Code, Cursor or ChatGPT. Repository tests run from CI on each push or pull request, from the command line, from webhooks or on a schedule. Slack carries notifications and does not start work. Sourcemomentic.ai/docs/mo/integrations/scripts-and-ci.mdread 2026-11-08
Model Flexibility & Routing	Partial
The agents draw on models from Anthropic, OpenAI, Google Vertex, xAI, Mistral, OpenRouter, AWS Bedrock and Azure, falling back to alternatives when one fails. Customers pin each agent to a version, which fixes its prompts and models together, and choose an effort level that sets how much work it does and multiplies the AI step rate, but they cannot pick a model directly or bring their own key. Sourcemomentic.ai/docs/ai/agents.mdread 2026-11-08
APIs / SDKs / MCP Extensibility	Full
A hosted MCP server signs in with OAuth 2.1 and offers tools to start, message, stop and list Mo sessions, upload builds, and list, triage and retest findings, and a local MCP server running on an API key adds 24 tools that create and edit tests and modules and run single steps. A REST API with bearer keys starts Mo sessions and pull request sessions and reads runs and run groups, and three command line tools cover Mo, web tests and mobile tests. Sourcemomentic.ai/docs/mo/chatgpt-and-claude.mdread 2026-11-08
Testing, Debugging & Optimization	Full
Momentic tests act as required checks in CI through their exit codes, and the Mo check passes, fails or stays neutral by set rules. Failures are classified as bugs, flakes or environment problems and routed to heal, warn or fail, quarantine rules act on pass and flake rates, and golden file visual diffs catch layout changes. A repair is delivered only once the original assertion passes again in a live browser, and Momentic scores its own AI steps on internal benchmarks. Sourcemomentic.ai/docs/reliability/auto-maintenance.mdread 2026-11-08
Browser / Computer-use	Full
Mo drives hosted Chromium browsers itself through Playwright, clicking, typing and navigating the app under test, and runs iOS simulators and Android emulators for mobile builds, each sub agent in its own browser or device. Repository tests use the same engines locally or hosted, including larger browsers for WebGL apps. Firefox, Safari and physical phones are not supported, so coverage stops at Chromium and simulated devices. Sourcemomentic.ai/docs/running-tests/hosted-test-environments.mdread 2026-11-08

The Agentic Index coverage score grades every vendor Full, Partial or Not documented against the same 14 buyer facing capabilities, from public evidence only. Each capability links to how all vendors in the index score on it. How this evidence is graded

Pricing

Free with $50 of usage a month, or Scale at $500 a month billed in advance as $500 of usage, with usage metered per step, minute and AI call.

Prepaid usage: test steps, AI steps, browser and emulator minutes and AI calls draw down a balance, with no seat charges; Scale prepays $500 a month and Enterprise is priced per test.

Free tier

Included quota

Free carries $50 of usage a month with five concurrent browsers, two Android and two iOS devices, 15 minute mobile sessions and 30 day result retention; Scale carries $500 of usage with 25 browsers, five Android and five iOS devices and five phone numbers.

What is public

Free and Scale prices, every usage rate, top up rules and plan limits, with Enterprise test based pricing by contact.

Billing mechanics

A prepaid balance drawn down by metered usage, with Scale prepaying $500 a month and top ups that never expire.

Cost watchouts

AI steps cost twice a plain step, and higher effort levels multiply that rate by up to three; iOS minutes cost nearly twice Android minutes, triage runs $1.25 a call, and Scale top ups stop at $1,000 a month above the base $500.

Variable cost rationale

Every plan is a balance drawn down by steps, AI steps, browser and emulator minutes and per call AI triage, and higher effort levels multiply the AI step rate, so monthly cost follows test volume and how much Mo is used.

Additional watchouts

Costs scale with steps, minutes and AI calls rather than seats, and Scale caps monthly top ups.

Overage / add-ons

There is no overage: runs pause when the balance reaches zero, and Scale customers add $50 to $1,000 at a time, up to $1,000 a month, in funds that never expire.

Sales call required

Mixed (some tiers require a call)

Free / trial

A free plan with $50 of usage each month and no card; there is no time limited trial.

Lowest paid plan

Scale at $500 a month, billed in advance as $500 of usage

Commercial notes

Momentic, Inc. sells self serve on Free and Scale and through sales on Enterprise, and has retired its older $125 pay as you go plan for new customers.

Key ambiguities

The pricing FAQ puts domain claiming on every plan while the docs limit it to paid plans, and pricing lists Google and SAML sign in while the security page names SAML and OIDC.

Cancellation / refund

Free and Scale are self serve with prepaid funds that never expire; Enterprise is an annual invoiced contract.

Support SLA / resale

Scale adds a shared Slack channel, and Enterprise adds an SLA, a dedicated account manager, SAML, SCIM, audit logs, one hour mobile sessions and higher API rate limits.

Missing data

Enterprise per test prices and contract terms.

Agentic Index verified 2026-11-08

Official pricing page Docs

Related vendors

Alternatives to Momentic

The closest documented capability profiles to Momentic among agent infrastructure platforms tracked by Agentic Index, ordered by similarity on the same 14 point evidence the rankings use. No vendor pays for placement.

OpenAI Agents platform12.5 / 14Fuller documented coverage on Deployment & Data Residency
Cloudflare13.0 / 14Fuller documented coverage on Deployment & Data Residency and Model Flexibility & Routing
Mastra14.0 / 14Fuller documented coverage on Deployment & Data Residency and Prebuilt Agents, Templates & Packs
Agno13.0 / 14Fuller documented coverage on Deployment & Data Residency and Prebuilt Agents, Templates & Packs
Microsoft Agent Framework12.0 / 14Fuller documented coverage on Deployment & Data Residency and Model Flexibility & Routing
Pipedream11.0 / 14Fuller documented coverage on Model Flexibility & Routing

Similarity is computed from each vendor's Agentic Index coverage score evidence, axis by axis, not from the totals. How this evidence is graded

Compare Momentic with other vendors →Compare all agent infrastructure platforms →See where Momentic ranks →AI agents for software development →

Data confidence: high