Contains affiliate links. We may earn a commission — it never changes the verdict. Details
The Stack / Automation
Manus
04 AutomationYes as a first general agent — Manus executing whole tasks (research reports, data work, working prototypes) from one brief is the closest thing to delegation in the consumer AI market. Credit burn varies wildly by task complexity, so treat the $20 Standard tier as an experiment budget, not a fixed cost.
The job in the stack
The autonomous-task seat: multi-step research, analysis, and build tasks executed end-to-end by an agent with its own computer.
Pricing
| Plan | Price | Notes |
|---|---|---|
| Free | $0 | 300 daily refresh credits; Chat Mode plus Manus 2.0 Lite in Agent Mode; 1 concurrent task, 2 scheduled automations |
| Standard | $20/mo | 4,000 credits/mo plus 300 daily refresh credits; Manus 2.0 Max / 2.0 / Lite; 20 concurrent tasks; annual saves 17% |
| Customizable | $40/mo | 8,000 credits/mo and up (credit tiers selectable); 7-day free trial; annual saves 17% |
| Extended | $200/mo | 40,000 credits/mo for heavy delegation; annual saves 17% |
| Team | from $20/seat/mo | All paid features plus SSO, training opt-out, team usage analytics, access control; annual saves 17% |
Pricing verified 2026-10-15. Plans change — the source of truth is Manus's own pricing page.
Models & versions
- Manus 2.0 (Max / 2.0 / Lite) Launched Sept 28, 2026 on the new Cascade agent harness. Free gets Lite only; paid plans get all three. Credits per task vary with complexity
- Cue Phone-number agent sold separately from membership; usage draws on a weekly Cue allowance, not Manus credits
The honest read
What works
- + Delivers finished artifacts — reports, spreadsheets, working pages — not chat answers
- + Watching its execution log teaches you what agents can actually do
- + Free tier's daily credits allow genuine evaluation before paying
What breaks
- − Credit consumption is unpredictable — a hard task costs 20x a simple one
- − Output quality still needs review; autonomous ≠ finished
- − Give the agent a leash: it browses and acts, so scope tasks deliberately
Use it for
- →Research reports and competitive analyses you'd otherwise assign out
- →Data collection and structuring tasks with clear success criteria
- →Operators learning what agent delegation feels like before building their own
Skip it if
- ×Your tasks are conversational — an assistant seat covers that cheaper
- ×You need deterministic, repeatable automation — that's n8n's job, not an agent's
Questions operators ask
?Manus or just using Claude?
Claude reasons with you; Manus goes away and does the task — browsing, collecting, building — then hands back the deliverable. If your bottleneck is thinking, Claude. If it's hours of execution, Manus. Many stacks now carry both.
What changed
- Re-verified on manus.im/pricing (rendered) and the help center. Prices hold: $20/4,000, $40/8,000+, $200/40,000 credits a month, annual −17%. New since July: Manus 2.0 (Sept 28; Max / 2.0 / Lite, Cascade harness), paid plans also get 300 daily refresh credits, a Team plan from $20/seat, and Cue phone agents sold separately. Dropped the Free 1,500-credit monthly cap note because the current help article no longer states it.
- Re-verified on manus.im/pricing: Standard $20 (4,000 credits), Customizable $40 (8,000), Extended $200 (40,000), 17% off annual, all unchanged. Paid tiers also list 300 daily refresh credits, plus 20 concurrent and 20 scheduled tasks. No verdict change.
- NEWS SCAN: Manus resumed fully independent operations on September 2, 2026, after China's NDRC ordered Meta's $2B acquisition (announced Dec 29, 2025) unwound in April 2026 over outbound-investment and tech-export rules. The Singapore-based founding team runs it again as an "independent agent lab." Separation involved deleting some user data generated on or after Dec 29, 2025 in specific regions to satisfy the regulatory unwind — worth knowing if you had a Manus account during the Meta window and haven't checked your task history since. No pricing or product change; added as a trust/ownership note.
- Freshness sweep: paid tiers unchanged ($20/4,000, $40/8,000, $200/40,000; annual −17%). Sharpened the free tier — Manus's help centre caps free daily-refresh credits at 1,500/month, so the 300/day figure is a ceiling, not a run rate. The public pricing page now foregrounds the paid tiers and lists "300 refresh credits everyday" as a paid feature, while the help centre still documents a Free Plan; treating the free tier as live.
- Freshness sweep: pricing re-confirmed unchanged (Free 300 daily credits, Standard $20/4,000, Customizable $40/8,000, Extended $200/40,000; annual −17%). Updated the model note — Manus now runs the 1.6 family (1.6 Lite / 1.6 / 1.6 Max).
- Released from staging. Re-confirmed — no changes: Free 300 daily credits, Standard $20/4,000, Customizable $40/8,000, Extended $200/40,000; annual −17%.
- Added to the stack (autonomous-task seat).
Give Manus a job
The autonomous-task seat: multi-step research, analysis, and build tasks executed end-to-end by an agent with its own computer.
Try Manus →Alternatives with different jobs
n8n
04 AutomationThe heavy-automation seat: complex, high-volume workflows and AI agent pipelines, self-hosted or cloud.
Claude
02 DecisionsThe thinking seat: long documents, hard reasoning, code, and anything where the words have to be right the first time.
Perplexity
01 IntelligenceThe research seat: current-events questions, competitive research, and any claim you need sourced before it ships.