← All toolsDOSSIER · Research assistant · SOLID · VERIFIED 2026-06-09

Manus67SolidBenchmark pendingGeneral-purpose autonomous research and task agentVerified 2026-06-09

Dossier · Research assistant

Manus

General-purpose autonomous research and task agent · last verified 2026-06-09

Solid

Research assistant

Manus

67/100

ROLEGeneral-purpose autonomous research and task agent

Editorial fit

Source quality

Citation honesty

Privacy controls

Value for money

Speed

FREEFree trial

$39/MOPlus

CUSTOMPro

SolidVerified 2026-06-09VP·METHOD

Manus flagship-ready dossier: General-purpose autonomous research and task agent.

What is Manus?

Manus is a tool in the VerdictPal Research assistant set: Cloud autonomous-agent platform for multi-step research, browsing, file work, reports, slides, websites, data analysis, scheduled tasks, and team/business workflows using credit-based plans. It scores 67 out of 100 on editorial fit.

Role: General-purpose autonomous research and task agentCategory: Research assistantEditorial · hands-on

Manus at a glance, with the date each field was checked.
Field	Value
Editorial fit	67 out of 100
Dossier status	Solid
Set	Research assistant
Role in a workflow	General-purpose autonomous research and task agent
Pricing	Free trial Free · Plus $39/mo · Pro Custom · Enterprise Custom
Training on your data	Not recorded as a plan-wide guarantee. Check the dossier's privacy notes and sources.
Last verified	2026-06-09

Editor's note

Manus belongs in VerdictPal as an agentic-workflow stress test, not as a magic intern. The tool should ask whether it saves work after verification, not whether it produces impressive artifacts.

Public facts · vendor & repo

List prices and pay-as-you-go entry points we can cite without running our own bench. Each tile links to a source when possible.

Research assistantSetVerdictPal tool ↗

synthesizedEvidenceQuality gate ↗

3Sources checkedSources ↗

2026-06-07Pricing checkedQuality gate ↗

Limits & product surface

Non-price vendor claims — multipliers, caps, and API scope. Detailed matrices live in subscription and SDK sections below.

Primary surface: consumer-appTool identity ↗
Modalities: text, code, dataTool identity ↗
Workflow roles: [object Object], [object Object], [object Object]VerdictPal editorial ↗
Alternatives tracked: ChatGPT, Claude, Perplexity, OpenAI Operator-style agents, GeminiVerdictPal comparison set ↗

How it works · agent loop

The public positioning for this product — the loop we score against on VerdictPal.

Start with the wedge

Exploratory multi-step tasks where a sourced report, spreadsheet, deck, or website is more useful than a chat answer

Run the representative task

Run a 5-task agent benchmark: literature scan, spreadsheet cleanup, source-backed report, slide deck, and web workflow; score autonomy, cost, audit trail, and corrections needed.

Check the failure modes

Credit consumption can be hard to forecast

Compare before recommending

Compare against ChatGPT, Claude, Perplexity before shipping advice.

Who it fits, where it fails

Best for

Exploratory multi-step tasks where a sourced report, spreadsheet, deck, or website is more useful than a chat answer
Users testing cloud agentic workflows before building their own automation stack
Low-stakes research/prototype work where credit burn and imperfect execution can be reviewed

Avoid if

You need predictable cost per task, exact reproducibility, or audit-perfect source trails
Files, client data, institutional research data, or regulated data cannot be processed by a cloud agent
You expect autonomous output to be publishable without source, calculation, and action review

Strengths

Multi-step cloud agent tasks
Web browsing and file automation
Reports/slides/websites/data analysis
Scheduled Tasks 2.0 for recurring work, web apps, and updated artifacts
300 daily refresh credits and 40,000 monthly credits in official pricing page
20 concurrent tasks
20 scheduled tasks
Team/business plans
Meta ownership note
Enterprise security positioning

Weaknesses

Credit consumption can be hard to forecast
Exact lower-tier plan prices remain unverified
Agent outputs can hide weak sources, failed steps, or bad calculations
Cloud execution exposes prompts/uploads/browser state/scheduled-task context/tool outputs
Meta ownership may alter buyer trust and data-risk analysis
Provider/data-sharing terms need re-check after acquisition

How it could improve

Institutional policy friction could be reduced with a local/on-prem deployment option, or at minimum with clear data-handling documentation that compliance teams can review without an NDA.
Opportunity cost scores 49/100, the weakest dimension on this dossier. Addressing this would meaningfully raise the editorial fit.
Privacy posture scores 44/100, the weakest dimension on this dossier. Addressing this would meaningfully raise the editorial fit.
Data portability scores 49/100, the weakest dimension on this dossier. Addressing this would meaningfully raise the editorial fit.

Search modes · 4 lenses

One input box, many retrieval postures. Filter by tier.

01Free

Default

Default path for Manus — verify limits on the live product.

02Pro

Deep work

Deep work path for Manus — verify limits on the live product.

03Pro

Focused task

Focused task path for Manus — verify limits on the live product.

04Free

Export & share

Export & share path for Manus — verify limits on the live product.

Competitive lens

Where Manus wins for cited research — and where a rival still belongs in the stack.

Manus is stronger when exploratory multi-step tasks where a sourced report, spreadsheet, deck, or website is more useful than a chat answer; ChatGPT may still win for narrower fit, procurement, or specialist depth.

Manus

Exploratory multi-step tasks where a sourced report, spreadsheet, deck, or website is more useful than a chat answer
Users testing cloud agentic workflows before building their own automation stack

ChatGPT

You need predictable cost per task, exact reproducibility, or audit-perfect source trails
Files, client data, institutional research data, or regulated data cannot be processed by a cloud agent

Scores and evidence

Metric lab · 16 dimensions

Click a tile for the editorial note. Color follows score: coral, yellow, mint.

Shape

Avg 68 · 44–100

67Editorial fit

Weighted roll-up across 13 dimensions for Research assistant; pending desk verification if rescored from public facts.

By group

Fit80

Editorial fit67
Wedge task fit82
Feature depth92

Cost56

Free-tier utility61
Cost-to-value58
Opportunity cost49

Trust72

Source grounding58
Privacy posture44
Failure transparency100
Evidence strength56
Transparency100

Workflow65

Integration reach74
Setup friction58
Reliability68
Competitive position75
Data portability49

Benchmark ledger

Public rows are vendor or third-party claims we logged with a date. Desk rows are reserved for VerdictPal self-run results.

Measure	Result	Source
Run a 5-task agent benchmark: literature scan, spreadsheet cleanup, source-backed report, slide deck, and web workflow; score autonomy, cost, audit trail, and corrections neededKeep recommendation locked until the named benchmark is run and reviewed.	Planned2026-06-09	VerdictPal benchmark plan

Editorial evidence · 1 entry

Task

Run a 5-task agent benchmark: literature scan, spreadsheet cleanup, source-backed report, slide deck, and web workflow; score autonomy, cost, audit trail, and corrections needed.

Desk trial 2026-07-05: personally tried; formal 5-task benchmark still pending.

↳ Keep recommendation locked until the named benchmark is run and reviewed.

This tool is in the VerdictPal Citation Fidelity protocol — frozen question bank, dual-reviewer grading, results still pending. Citation Fidelity v0.1 →

Benchmark coverage · 1 benchmark

Citation Fidelity (VerdictPal Lab)

Benchmarks that evaluate or report on this tool. Scores and caveats live on the benchmark page.

Test scenarios · hands-on lab

How we exercised the product. Step through each run before you trust the scores.

Test run

Step 1 of 4

Start from the persona in best-for item 1.

Research log · desk notes

What we learned while building this dossier — not vendor copy.

2026-06-04Notion Tool pipeline (Card pipeline) ↗

Notion research pass

Imported Notion state-of-art fields and corrected Scira-gate status to solid.

2026-06-09VerdictPal git ↗

Flagship-ready structure

Converted imported research into metrics, panes, scenarios, benchmark rows, comparison notes, and pricing deck.

What it costs, what it keeps

Pricing deck · checked 2026-06-09

Desk pick

Plus

$39

month

Verify credits on pricing page
Verify live regional checkout before publishing procurement advice.

Pro

Custom

month

Higher limits
Verify live regional checkout before publishing procurement advice.

RAW · checked 2026-06-09: Verified 2026-06-07 from official Manus pricing/product pages: credit-based plans with daily refresh credits; pricing page lists 300 refresh credits every day, 40,000 credits/month, in-depth research for large-scale tasks, professional websites with data analytics, slides for batch production, Wide Research for heavy use, early beta access, 20 concurrent tasks, 20 scheduled tasks, team/business plans, and enterprise security/compliance positioning. Exact lower-tier plan names/prices still need live checkout/full-table verification.

Privacy deep-dive · checked 2026-06-09

Official Manus privacy snippet says policy last updated 2026-04-30 and names Butterfly Effect Pte. Ltd. as provider of Manus AI platform, mobile apps, and browser extensions. Official Manus blog says Manus is now part of Meta, will keep selling/operating subscriptions through its app and website, and will continue operating from Singapore. Cloud agent tasks mean prompts, files, browser/app context, scheduled-task state, artifacts, and outputs are vendor-processed. Re-check DPA, retention, training, enterprise security, and Meta/Manus data-sharing terms before sensitive use.

Under the hood

Vendors named on the product about page — useful for procurement and privacy reviews.

multi-step research tasks: Manus
browser and file automation: ChatGPT
report drafting: Claude

Alternatives and context

Alternatives

ChatGPT Claude PerplexityOpenAI Operator-style agentsGemini

Head-to-head

Manus vs ChatGPT Manus vs Claude Manus vs Perplexity Manus vs Gemini

Compare side-by-side → open compare

Workflow roles

How this tool fits into a composed research stack:

multi-step research tasksbrowser and file automationreport drafting

Deep panes · 6 lenses

Editorial lenses only. Subscription and API pricing live in their own sections above.

Lens 01

Positioning

Manus is framed here as General-purpose autonomous research and task agent. Manus is an autonomous-agent workflow tool, but it needs strict caveats. Official pages now show Manus as part of Meta, while the privacy page snippet still names Butterfly Effect Pte. Ltd. and was last updated 2026-04-30. Pricing details were only partially extractable from official pages, so exact plan comparisons should remain guarded.

Checked: 2026-06-07
Evidence: synthesized

Exploratory multi-step tasks where a sourced report, spreadsheet, deck, or website is more useful than a chat answer
Users testing cloud agentic workflows before building their own automation stack
Low-stakes research/prototype work where credit burn and imperfect execution can be reviewed

Sources and provenance

Source ledger · 3 checked

Product sitechecked 2026-05-30
Pricingchecked 2026-05-30
Privacychecked 2026-05-30

Quality gate · Solid

Evidence ready

Benchmark pending

Verdict history

2026-06-04Imported Notion state-of-art fields and corrected Scira-gate status to solid.
2026-06-09Converted imported Notion research into a full flagship-ready dossier template with metrics, panes, pricing deck, scenarios, benchmark rows, and comparison slices.

Related dossiers

General AI assistant

ChatGPT

68/100

ROLEGeneral-purpose assistant baseline and multimodal draft critic

OpenAI’s general-purpose assistant for drafting, analysis, multimodal exploration, file work, coding, agents, and baseline AI comparison, with separate consumer, Business/Enterprise, and API trust boundaries.

Watch outYou need API-only usage and are evaluating API economics, not the chat subscription

Editorial · hands-onSolid

General-purpose drafting, critique,Students who want

ROLELong-context drafting and analysis desk

Anthropic’s assistant for careful drafting, long documents, Projects, Artifacts, Research, web search, memory, Claude Code, Cowork, connectors, and API workflows, with separate consumer, team, enterprise, and API trust boundaries.

Watch outFootnotes and domain audits are the deliverable — use Perplexity, Kagi, Scira, or another cited-search workflow first

SynthesizedSolid

Thesis chapters, ethics-heavyProjects that keep

ROLECited search front door and web-grounded API platform

Cited answer engine and AI search platform with consumer Pro/Max search, Deep Research, Spaces/files, and developer APIs including Sonar, Search, Agent, and Embeddings.

Watch outYou plan to paste the answer into a graded bibliography without opening every citation

Editorial · hands-onFlagship

First-pass source mapsPolicy, news, product,

ROLEGoogle ecosystem multimodal assistant

Google’s multimodal assistant and AI plan ecosystem, with Search grounding, Deep Research, Deep Search in AI Mode, Gems, Live, image/music/video generation, NotebookLM benefits, Google storage bundles, and a separate Gemini API.

Watch outYou need numbered footnotes and domain audits like a cited search engine

Editorial · hands-onFlagship

Google-heavy users whoDrafting rewriting Google

From$7.99/MOGoogle AI Plus

ROLECited academic Q&A and claim-level literature scanning

AI academic search engine for cited answers, claim-level literature Q&A, Research Agent workflows, paper analysis, and deeper review-style summaries over a large peer-reviewed research corpus.

Watch outYou need a full systematic-review workflow with reproducible search strings, screening audit trails, and PRISMA documentation

SynthesizedSolid

Fast yes/no orStudents who need

From$15/MOPro (monthly)

ROLEPaper discovery, evidence-table extraction, and systematic-review support

AI research assistant for searching, summarizing, chatting with papers, screening studies, extracting fields, generating research reports, and supporting systematic reviews across large scientific corpora and uploaded documents.

Watch outYou need a cheap casual search tool rather than a high-cost research workflow

SynthesizedSolid

Building structured evidenceEarly-stage systematic review

Questions this dossier answers

Every answer below is assembled from the dated fields on this page. Nothing is written separately for search.

Is Manus worth using?

Manus scores 67 out of 100 on editorial fit and carries Solid dossier status. Its job in a research workflow is: General-purpose autonomous research and task agent.

Who is Manus best for?

Manus earns its place when you need:

Exploratory multi-step tasks where a sourced report, spreadsheet, deck, or website is more useful than a chat answer
Users testing cloud agentic workflows before building their own automation stack
Low-stakes research/prototype work where credit burn and imperfect execution can be reviewed

When should you not use Manus?

Skip Manus in these cases:

You need predictable cost per task, exact reproducibility, or audit-perfect source trails
Files, client data, institutional research data, or regulated data cannot be processed by a cloud agent
You expect autonomous output to be publishable without source, calculation, and action review

What does Manus cost?

Free trial Free · Plus $39/mo · Pro Custom · Enterprise Custom. Pricing last checked 2026-06-09.

Does Manus train on your data?

Not recorded as a plan-wide guarantee. Check the dossier's privacy notes and sources.. Privacy terms last checked 2026-06-09.

What goes wrong with Manus?

The dossier publishes 6 failure modes for Manus, and they stay published whether or not the vendor likes them:

Credit consumption can be hard to forecast
Exact lower-tier plan prices remain unverified
Agent outputs can hide weak sources, failed steps, or bad calculations
Cloud execution exposes prompts/uploads/browser state/scheduled-task context/tool outputs
Meta ownership may alter buyer trust and data-risk analysis
Provider/data-sharing terms need re-check after acquisition

What are the alternatives to Manus?

The closest options to Manus are ChatGPT, Claude, Perplexity, OpenAI Operator-style agents, Gemini. Each one that has a tool in the atlas is linked from this dossier, with a head-to-head comparison.

The Pack · Editorial newsletter

New tools in your inbox. Free.

One short email when a tool ships or changes status. No tracking, no third-party analytics. Unsubscribe in one click.