Maze Review, Pricing & Alternatives (2026)

Maze is a product research platform for product, design, and research teams. Founded in 2018 by Jonathan Widawski and Thomas Mary, the London based company was formerly known as Maze.Design, Inc. Maze says more than 60,000 teams use the service, and its panel covers more than 5 million participants worldwide. Participants come from more than 130 countries and can be filtered across more than 400 demographic attributes. The platform supports prototype, website, mobile, survey, interview, card sorting, and tree testing workflows. Its audience includes designers, product managers, researchers, and product leaders who need evidence during product development. Maze is active, available on the web, and offers an API for teams that need connected workflows.

A typical Maze study begins with a research goal, a prototype, a website, an app, or a question set. The builder uses blocks for tasks, surveys, conditional logic, screening, card sorting, tree testing, five second tests, and ratings. Teams can import prototypes from Figma, Adobe XD, and Sketch, then share a link or recruit through Maze Panel. Interview Studies add scheduling, recordings, transcripts, observers, and highlight reels on eligible plans. Reports combine quantitative results with audio, video, text responses, clips, filters, exports, and shareable findings. Teams can also recruit their own customers through links, databases, or contextual prompts. This broad method coverage supports recurring mixed method research without requiring separate tools for every study type or reporting need.

Maze offers AI study creation, AI Moderator, AI conversation blocks, and Maze MCP, although it does not identify one underlying model for its AI features. In July 2026, the AI conversation block added adaptive follow up questions to unmoderated studies, connecting observed behavior with participant explanations. An August 2026 release opened MCP to Enterprise teams and added screener result blocks, element click reporting, usage dashboards, and up to eight AI Moderator learning goals. Maze’s 2025 releases included observer rooms, transcripts in more than 58 languages, governance controls, highlight reels, an AI Moderator beta, and a mobile testing app. Teams should review Maze’s privacy policy, data processing terms, security documentation, and AI terms before uploading sensitive research or enabling these features.

Maze has a Free plan costing $0/month, intended for individuals with limited features after the trial. New accounts receive premium access for 30 days or three studies, with no credit card required. Paid pricing is custom or shown at checkout, depending on seats, studies, features, and panel usage. Costs can vary by study allowances, advanced features, AI access, moderated research, governance, and panel credits. Enterprise access includes negotiated features and usage. Buyers should confirm monthly or annual billing, included studies, seats, credits, security controls, governance, AI access, participant permissions, data retention, and overage terms before purchase. Maze is less suitable for buyers needing published prices for every paid plan, a self hosted deployment, or one simple survey without research operations.

Maze suits teams running recurring prototype and usability studies, combining surveys, interviews, and behavioral testing, or recruiting participants through targeted panel filters. Lyssna is a simpler option for prototype and survey testing. UserTesting offers a large contributor network and recorded feedback, while Lookback focuses more closely on interviews and usability sessions. Maze combines testing, recruitment, analysis, reporting, moderated research, unmoderated research, and AI assistance in one workspace, but panel recruitment and advanced features can add usage costs. Choose Maze over Lyssna when mixed methods and recurring studies matter, or over UserTesting when one workspace matters more than recorded contributor feedback. Pick Maze if your team wants recruitment, AI assistance, and reporting together.