Merlin is a cross-platform AI assistant for AI Office and Efficiency. The core selling point is not a single chat window, but the integration of multi-model Q&A, web page summary, file Q&A Projects knowledge base and in-browser writing portal into one account system.

Merlin - In-depth tool analysis

Core parameters and statistics

Parameters	Current public information
Tool type determination	Productivity/Business-side application (Rule D)
Reason for determination	The main delivery form is a browser extension for Web and mobile efficiency tools, rather than an independent model API; what users purchase is a finished workbench, not the underlying inference service
Official positioning	All-in-One AI extension to Write, Summarize, Code & Play
Core entrance	Official website Chrome extension Web Chat, iOS, Android
Platform coverage	Web, Android, iOS, and Chrome Web Store pages also state that the same account can be used across Chrome, Edge, Windows, and Mac
User scale	Official website public Trusted by 20M+ users
Extension Ratings	Chrome Web Store 4.8/5, 8,794 ratings, 900,000 users
Mobile port reputation	App Store 4.6/5 (5,685 Ratings); Google Play 4.1/5 (12,800 reviews, 500,000+ downloads)
Free tier	Official website FAQ public 102 free inquiries per day
Paid starting point	Pro annual payment is $19/month; Teams is $15/seat/month, minimum 5 people

Brief review in one sentence: Merlin is not another chat box, but an AI workbench that integrates the broken actions of "reading, searching, writing, modifying, and organizing files on the web page" into a unified entrance.

Publicity verification: The official website puts "multiple models in one", "no need to cut tabs" and "one account across devices" at the top. This selling point is basically established, because the public page also provides browser sidebar, search enhancement, web/video/PDF summary Projects knowledge base and mobile terminal synchronization entrance. The real pain point it hits is not the model itself, but the switching cost in high-frequency information processing tasks.

User and market recognition

Public recognition signals: The official website directly lists 20M+ users; Chrome Web Store shows 4.8 points and 8,794 ratings from 900,000 users; App Store shows 4.6 points and 5,685 ratings; Google Play shows 4.1 points and 500,000+ downloads. These numbers don't prove that it's suitable for every industry, but they are enough to show that Merlin has passed the "niche experimental tool" stage.

The true meaning of corporate endorsement: The homepage of the official website displays brand logos such as Walmart, Cisco, Netflix, Pinterest, Zoom, Sony, eBay, and Uber. A more reasonable interpretation here is "market endorsement signal at the brand level" rather than directly inferring that it is standardized procurement by large-scale enterprises. For enterprise buyers, what should really be verified is SSO, permissions, auditing and contract terms, rather than just looking at the logo wall.

Business implications of word-of-mouth structure: In public reviews, usage experiences such as "research", "summary", "less switching of applications" and "Cmd/Ctrl+M shortcut keys" appear frequently, indicating that Merlin's strongest point is not a single point model, but the integration of AI touch points into the user's original workflow. On the other hand, if the team does not work frequently in the browser and document environment, this advantage will be significantly reduced.

Cost advantage

C-side/Individual: The free tier has 102 queries per day, which is the most direct threshold design for Merlin to attract new users. But free does not mean you can use it casually. Advanced models will consume credits at a higher rate. The FAQ on the official website clearly states that models such as GPT-4, Gemini 1.5, and Mistral Large will consume 30 queries at a time, while GPT-3.5 and Gemini basic file Claude 3 Haiku will only consume 1 query. For individual users, the free tier is more like a "trial workflow" rather than "long-term high-intensity white prostitution."

Developer/API Alternative Perspective: The Merlin pricing page clearly states that it does not provide a direct API and that the product goal is a "multi-model user-friendly interface". This means that it is suitable as a terminal productivity tool, not as a programmatic orchestration base. If the development team wants to connect multi-model calls to its own system, it still needs to purchase APIs, perform authentication and billing management separately.

Enterprise/Team: Teams is priced at $15/seat/month annually, with a minimum of 5 people, and includes full Pro capabilities, admin and access control, usage dashboards, larger context windows, and security representations compliant with SOC2, GDPR, and ISO 27001. The explicit cost is not high, but the implicit cost comes from permission boundaries, data access scope, employee training and prompt word management.

Quantitative deduction of cost reduction and efficiency improvement (unofficial commitment): For positions such as marketing, consulting, and operations, which are mainly "browsing web pages, summarizing information, and writing outbound texts", Merlin is more like pressing a round of tasks from "opening 5 to 8 pages and manually moving them" to "complete the first draft and summary within one browser session." It is reasonable to manually shorten a competitive product speed reading or a cold start email draft from 45 to 60 minutes to 15 to 25 minutes; but this is only true for high-frequency, formatting tasks, and not for scenarios that require in-depth original judgment.

Main functions

• In-webpage AI sidebar: Chrome extension supports Ctrl/Cmd + M to call out the assistant to directly answer questions, rewrite and summarize in the context of web pages, search results, social media and emails. The value of this capability is not "being able to chat", but embedding the call entry into the current page to reduce copying and pasting.
• Unified access to multiple models: Both the official website and the extension page clearly list GPT, Claude, Gemini, DeepSeek, Mistral, Llama, Grok and other models. The real meaning is to reduce multiple subscriptions and multiple account switching, but model quality and quota are still subject to Merlin's own fair use restrictions.
• Projects Knowledge Base: Pricing page description Projects can maintain a knowledge base and have conversations through Google Drive, linked YouTube videos and other sources. Pro and Teams give a stronger RAG. Its hidden linkage lies in "data accumulation + follow-up questions", which has more long-term value than a one-time summary.
• Crafts Generation Workshop: The official website includes infographics, interactive charts, React + Shadcn application fragments, and web component generation in Crafts. What is more important to the business side is not "generating diagrams", but that the research content can be directly converted into displayable products, reducing the secondary transfer from text to charts to light prototypes.
• File and Media Q&A: The extension page publicly supports website PDFs, documents, pictures, YouTube video summaries and Q&A, and the official website also mentions that PDFs can be uploaded as context. Its hidden linkage is to combine "read files", "ask models" and "export summaries" into one step, which is especially suitable for course materials, pre-sales materials and project reviews.

Expert view: Merlin's functional synergy is not a simple stack, but a short link: first capture the context in a web page or file, then sink the context into Projects, and finally use Crafts or a writing portal to directly generate emails, charts, social media copy, or code snippets. The real savings are in the cost of cross-software switching and context reconstruction, rather than the cost of a single generated token.

Model and version evolution

Version form judgment: Merlin is not a typical semantic version SaaS, it is closer to a "multi-terminal product collection with continuous iteration". Therefore, you can’t just look at the homepage of the official website, but look at whether the three public nodes of expanding iOS and Android are advancing simultaneously.

Merlin’s latest public node

Node	Date	Public Signal	Business Meaning
Chrome extension 8.0.11	2026-06-09	Chrome Web Store details page is visible	The browser entrance is still the core distribution aspect of Merlin, and the language has been expanded to 55
Android update node	2026-01-08	New changes in Google Play indicate faster attachment uploads and bug fixes	Note that file upload and mobile usability are still the focus of continued polishing
iOS 6.1.6	~2026-01	App Store “What’s New” display fixes chat share bug	The mobile terminal focuses on stability and sharing links, rather than pursuing complete isomorphism with the browser terminal

Evolution Main Line: Merlin first relied on the in-browser AI assistant, and now it has obviously expanded to "multi-model entrance + knowledge base + lightweight analysis and creation workbench". Projects, Analyst Mode, Tasks, MCP Server Integrations, and Deep Research Mode appear on the pricing page, indicating that it is trying to upgrade from a "browser plug-in" to a "personal workbench".

Border Reminder: These capabilities appear on the public pricing page and do not mean that all entrance experiences are exactly the same. In particular, there are likely to be differences in the functional maturity of the browser extension Web Chat and the mobile terminal, and the selection should be verified one by one according to the actual usage entrance.

Technical advantages

Mechanism 1: The context is obtained locally instead of manually moved. Merlin treats web pages, search results, video PDFs, and email input boxes as contextual entry points, reducing the need to copy content into the chat box and then ask questions. The effect is faster drafting and faster summarization; suitable for information workflows dominated by browsers.

Mechanism 2: Unified orchestration of multiple models, but using product strategies to encapsulate complexity. Both the official website and pricing page emphasize multi-model access and Merlin Magic automatic model selection. For ordinary users, this lowers the threshold for model selection; but the price is that you do not have the fine control of the underlying API, nor can you decide the routing logic yourself like the engineering team.

Mechanism 3: Projects + Crafts form a "material to product" relationship. One is responsible for precipitating context, and the other is responsible for generating diagrams, light applications, code and exhibits. This feature has more product barriers than the chat function alone because it shortens the link from information collection to deliverable shaping.

Mechanism 4: Cross-device account synchronization. The official website emphasizes syncing history, prompt words, and chatbots across all devices. The advantage is that the mobile terminal supplements information and the desktop terminal completes the output; the disadvantage is that if the team does not have clear data attribution rules, personal and team information will be mixed in the same usage habits.

Compliance and Security Disclosure Information: The official website and pricing page both mention GDPR, SOC2, and ISO 27001; the Google Play data security page states that data will not be shared with third parties, that data transmission is encrypted and deletion can be requested; the Chrome Web Store lists that it will process personally identifiable information and location data, and states that it will not sell data for purposes unrelated to core functions. The basic security posture can be seen, but more detailed log retention, model-side data processing and enterprise isolation strategies still need to be further confirmed by official sales or terms.

How to use

Entrance 1: Browser high-frequency tasks. Install the Chrome extension and pin it to the toolbar, use Ctrl/Cmd + M as the main entry, and instantly call summary, translation, rewriting and Q&A when searching, email, social media, and web page reading. This is the easiest way to use Merlin.

Entrance 2: Web Chat and Projects. Add frequently used links, Google Drive files, or YouTube content to Projects and keep asking. This entrance is more suitable for tasks that require context reuse, such as special research, training material Q&A, and pre-sales FAQ.

Entrance 3: Crafts and Analyst Mode. Enter Crafts or Analyst Mode when the task goal is not just to "know the answer" but to deliver charts, infographics, web snippets, and data analysis results. They are more like output layers than a universal starting point.

Recommended implementation path: First use a high-frequency task pilot, such as "weekly report data collection" or "sales email first draft"; after confirming the summary quality, citation stability, quota consumption and manual review costs, then expand to team shared knowledge base and multi-role collaboration.

Human-machine collaboration boundary: Web page summaries, email drafts, social media response suggestions, and video summaries can be highly automated; when it comes to formal contracts, external announcements, irreversible customer commitments, financial digital explanations, and high-value consulting delivery, manual final confirmation points must be retained.

Product Pricing

Plans	Current Public Prices	Who Is Suitable for	Key Limitations
Free	$0	Individual users who want to experience the in-browser workflow first	102 free queries per day, premium models are consumed at higher rates; many features have limited access
Pro	$19/month (annual payment, annual bill $228)	High-frequency research, writing, summarizing and multi-model switching users	Still subject to fair use restrictions; advanced models, image models Deep Research, Projects, etc. are limited by monthly or functional quota
Teams	$15/seat/month (annual payment, minimum 5 people)	Teams that want to unify member access and management	Requires at least 5 people, more permissions and a larger context window, but the terms and conditions should still be subject to the official sales page

The truth about the free version: Merlin is not "unlimited free multiple models", but "stratified with query quota and function limit". If a team relies primarily on GPT-4 level model Deep Research or image generation, the free version will quickly hit a wall.

The real selling point of Pro: Not just one more model, but packaged with extended Projects, Crafts, image generation, in-Chrome writing, cross-device synchronization, and a larger contextual window. For users who work in a browser every day, $19/month may be more cost-effective than buying a chat product alone; for people who only ask a few questions occasionally, it may not be worth it.

Standing for the official real-time page: Merlin's model quota and image quota are frequently updated, and differential restrictions such as OpenAI o3/o1, Gemini 2.5 Pro, DeepSeek R1, Grok 3, FLUX, Ideogram, GPT Image 1, etc. have appeared on the pricing page. You should verify on the real-time page before purchasing and should not rely on historical screenshots or old articles.

Application scenarios

• Research and Competitive Product Speed ​​Reading: When switching data back and forth between web PDF, YouTube and search results, Merlin is the most stressful because it can integrate search, reading, questioning, and writing into one browser session.
• Marketing and Content Operations: Google search Companion, Gmail writing
• Education and Knowledge Work: Course handouts, PDF summaries of paper materials, reference writing and study aids are typical usage areas repeatedly emphasized on the official website.
• Pre-sales and consulting work: Projects are used to organize data sources, and Crafts are used to convert key points into charts or light displays. They are suitable for preliminary research and internal preparation, but are not suitable for handing over the results directly to customers without review.
• Lightweight data and prototype work: Analyst Mode, Crafts and code snippet generation can cover part of the table analysis and front-end prototype needs, but it is still "fast enough to produce the first version" and is not a complete BI or formal development replacement.

Dimensionality reduction attack scenario: There are many sources of information, the task pace is fast, and the delivery is "first draft first and then iterate". It is often best to use Merlin.

Applicable people

• Students and research-oriented individual users: People who need to quickly summarize documents, videos, web pages, and repeatedly ask questions can obviously benefit from Projects and multi-model switching.
• Marketing, Operations, Growth and Content Team: People who frequently operate in Gmail, LinkedIn, X, and Google search can best reap the dividends of in-browser portals.
• Consultants, Product Managers, Analysts: People who need to quickly organize data from multiple sources into outlines, charts, and first-version insights will find the Projects + Crafts combination less switching than a single chat product.
• Lightweight Creators and Independent Developers: People who want to quickly create copywriting, web page snippets, infographics and light prototypes can use it as a production support desk.

Dissuade/not applicable to people:

• Engineering teams who need low-level API control: Merlin explicitly does not provide a direct API and is not suitable as a programmatic calling base.
• Formal Delivery Process for a Strong Compliance Industry: If content must be rigorously audited, traced, versioned, and permissioned isolated, Merlin's public capability description alone is not enough.
• In-depth long text originality or high-precision professional analysis scenarios: Merlin is good at accelerating first drafts and abstracts, but it does not mean that it can replace final professional judgment.
• People who almost never work in the browser: One of Merlin's big moats is the browser context. If the main work occurs in the IDE, local BI, design software or corporate intranet system, the value of this layer will be discounted.

Summary and Outlook

Core positioning: The most worthwhile thing about Merlin is not that "one account uses many models", but that it embeds multiple model calls into browsers, mobile terminals and knowledge base links, making information intake and first draft generation more consistent. For people who frequently research, summarize, rewrite, and post text, the value of this reduced switching is more immediate than the model list itself.

Current Limitations: The free tier query and premium model magnification limits are very clear; Pro and Teams are still subject to the fair use policy; no direct API is provided; the capability maturity of different ends may not be consistent; enterprise-level permissions, data isolation and audit granularity still need to be confirmed according to real-time terms.

Not suitable for boundaries: If the team only wants the strongest native single-model experience, requires extremely deep workflow automation, or all high-value results must be strictly accountable, Merlin is not necessarily the best solution. It is more suitable for the efficiency improvement of "pulling a 60-point fragmented process to an 80-point deliverable" rather than taking on the final responsibility of 100 points.

Procurement/Adoption Risk Assessment: Merlin’s main procurement risk is not in the surface unit price, but in three points. First, model and function quotas are updated quickly, and the budget must be reviewed on a real-time page; second, if the team regards it as a "final output tool that does not require review", errors will be amplified; third, when it comes to sharing the enterprise knowledge base with members, it is necessary to first verify the fair use restrictions in data boundaries, administrator visibility, and terms. A more stable way to adopt it is to first let 3 to 5 high-frequency positions run a two-week pilot, quantify "how much switching, how much first draft time, how much rework is saved", and then decide whether to expand the coverage.

Related tools: Notion AI, google-workspace

Version Info

• Chrome extension public version ：The public version of Chrome Web Store has a page display extension size of 4.11 MiB, supports 55 languages, and continues the in-webpage summary, writing and multi-model sidebar capabilities. （2026-06-09）
• Android mobile updates ：The Google Play page shows that the version update of this date includes faster attachment uploading and bug fixes, indicating that Merlin has listed file uploading and mobile usage as the focus of continuous optimization. （2026-01-08）
• iOS 6.1.6 ：The App Store page displays Version 6.1.6, and notes that the chat sharing problem has been fixed; there is no official precise year display on the page, and the current public page approximates it as January 2026. （~2026-01）