Merlin

Summary Merlin is an all-in-one AI assistant from Foyer Tech built around a web app and browser extension. It combines multiple upstream models with page-aware chat, document and video summaries, Projects, Deep Research, spreadsheet analysis, image tools and lightweight coding/diagram creation. The exact Chrome extension currently shows 4.8/5 from about 8,800 ratings and 900,000 users, giving the established browser workflow meaningful product-level validation. The harder questions are economics and reliability: Pro is advertised at $19/month on the current comparison surface, but usage remains subject to fair-use and internal cost ceilings, while a small Trustpilot population is heavily negative on credits, performance, support and refunds. Rating: Coming soon.

Merlin's current homepage positions the product as a persistent AI layer over the browser: summarize, search, repurpose and create from a webpage without switching tabs, ask questions about PDFs and videos, draft email and social content, and invoke the assistant with Ctrl/Cmd+M.¹ The current Chrome Web Store confirms that this is a mature distribution surface rather than a demo: version 8.2.3 was updated August 18, 2026 and has roughly 900,000 users.²

The extension is especially useful when context already sits in the page the user is viewing. That reduces copying between sites and separate chat products. Chrome describes Merlin as supporting website/video/document research, live web and file context, email/social assistance and more than 70 tools.²

The current pricing and product pages expose models from OpenAI, Anthropic, Google, DeepSeek and other providers, with Merlin Magic selecting models automatically for some use cases.³¹ Current plan comparisons include Claude Opus/Sonnet, GPT-family models and many image/video models, but the catalog changes with upstream availability.

Those model names do not make upstream benchmark scores Merlin scores. The product adds routing, context handling, retrieval, browser integration, Projects, file processing and usage controls around them. A model that benchmarks well directly can still behave differently when limits, context packing or product orchestration change.

Projects can hold reusable knowledge bases with sources such as Google Drive, links and YouTube content, while Deep Research adds web browsing and reasoning. Analyst Mode is designed to analyze spreadsheets, write analysis code and visualize results.³ Crafts adds diagrams and React/Shadcn-style app snippets from prompts, and the product also exposes image generation/editing and specialist tools.¹³

This breadth is valuable for generalists but creates an evaluation problem: extension summarization, spreadsheet analysis, deep research and image generation have very different failure modes. There is no strong current independent controlled benchmark showing that Merlin as a whole outperforms the native model products on those tasks.

Merlin advertises web, browser-extension, iOS, Android, Windows and macOS access with synchronized history.¹ Exact mobile-store verification is less clean than Chrome in this review cycle because a prominent same-name Apple result belonged to a different developer. It is therefore excluded rather than silently attributed to Foyer Tech.

Merlin currently offers Free, Pro and Teams.³ The current comparison page displays Pro at $19/month. A localized annual surface describes Pro as billed at $228 for a year, consistent with a $19 monthly equivalent. Teams is seat-based with a minimum team size of five and adds administration, usage dashboards and stated SOC 2, GDPR and ISO 27001 controls.³

Free provides restricted access to core and higher-tier models, limited extension features and limited tools. Pro materially expands model, image, research, extension and analysis usage rather than unlocking a completely different product.³

The most important pricing caveat is fair use. Merlin's Terms say excessive usage can trigger temporary suspension and define internal service-cost thresholds: standard paid accounts may be stopped for the rest of a month if calculated usage costs exceed $100, and for the rest of a day above $16; discounted plans may receive proportionally lower ceilings.⁴

The pricing page likewise says fair-use limits are proportional to the price paid after discounts, and current Pro documentation describes higher multiples of Free rather than a promise of unconstrained frontier-model use.³ Users who buy Merlin to replace several heavy native subscriptions should therefore test their actual workload before treating the $19 price as equivalent to unlimited access to every included model.

The Terms also allow suspension for usage whose internal costs reach 2.5× the monthly fee and say outlier usage can be restricted at the company's discretion.⁴ These limits are economically understandable for an aggregator, but they reduce predictability.

The current pricing FAQ says Merlin does not provide direct API access and is designed as a user-facing interface to multiple models.³ That makes it a productivity product rather than a clean infrastructure substitute for teams that need programmable model access.

The exact Chrome Web Store listing currently reports 4.8/5 from about 8,800 ratings and roughly 900,000 users.² For a browser-centric product, that is the most relevant public satisfaction signal available in this cycle. Chrome also marks Foyer Tech as Featured and following recommended extension practices.

The store says the extension handles personally identifiable information and location. The developer declares that data is not sold to third parties, not used for unrelated purposes and not used for creditworthiness decisions.²

Product Hunt supplies smaller but more detailed qualitative evidence. Current user comments praise Projects, preserved context, comparing multiple models and the browser workflow, while complaints include free-use reductions, glitches and the absence of Firefox support in older feedback.⁵ These comments are useful context, not a representative rating population.

The current getmerlin.in Trustpilot profile is 2.0/5 from 26 reviews, with 73% one-star and 19 reviews in the preceding 12 months.⁶ Recent complaints include rapid credit consumption, model quality, slow or incomplete responses, failed subscription activation, support delays and refund disputes. Some company replies acknowledge earlier OCR/document-processing or performance problems and say later updates addressed them.

The sample is tiny and strongly self-selected, so it should not be treated as a direct measure of overall product satisfaction. It is nevertheless relevant because the complaint themes align with the product's most consequential buyer risks: usage economics, reliability and post-purchase support.

The current privacy policy took effect January 20, 2026 and covers web, browser extensions, mobile apps, the AI Note Taker and API-related services.⁷ It says Merlin collects prompts, uploaded files, conversation history, usage information and—when the extension is enabled—browsing Activity Data associated with visited pages.⁷

The policy states that collected information can be used to develop, test and improve AI models and services and that aggregated usage patterns may be used to refine AI models. It also says Activity Data can be used to provide page-aware features, security, troubleshooting and extension-performance improvement.⁷ That is broader than a strict no-training/no-improvement posture and matters for confidential work.

Retention is more explicit: user content is retained for 30 days after a deletion request unless law requires longer; usage data can be retained for up to two years; raw extension browsing logs are retained as necessary for stated purposes, with shorter raw-log periods where feasible and aggregated/de-identified analytics retained afterward.⁷

The same policy says users retain ownership of their inputs and Merlin does not claim ownership of AI outputs. It also describes encryption in transit and at rest, access controls and regular security assessments.⁷

Merlin's Terms provide the product as-is and disclaim warranties around accuracy, reliability, availability and security.⁴ Liability is capped at amounts paid in the six months preceding a claim. The user remains responsible for submitted content and for verifying outputs.

For professional use, the combination of broad upstream-model access, browser Activity Data and variable fair-use limits means procurement should evaluate privacy, data-processing and usage economics together rather than looking only at the model list.

Merlin fits browser-heavy knowledge workers who regularly move between search, webpages, YouTube, PDFs, email and writing and want one shortcut-driven assistant instead of several separate tabs.¹²

It is also attractive to users who actively compare models. One subscription can expose several providers plus Projects, research and creation tools without separate consumer accounts for each.³

Students, marketers, founders, consultants and analysts are explicit product targets. Projects and document/video summaries are particularly relevant when the same source set is reused repeatedly, while Analyst Mode targets spreadsheet-heavy work.¹³

Merlin is less suitable for users who need guaranteed high-volume frontier-model access, predictable fixed quotas or direct API integration. The fair-use ceilings and credit/top-up mechanics can make a native provider subscription easier to reason about for sustained heavy workloads.⁴³

Privacy-sensitive teams should also distinguish ordinary browser assistance from confidential client or internal work. The current policy allows collection of user content and extension Activity Data and permits service/model improvement uses.⁷