Manus
Manus is a genuinely impressive autonomous AI agent that can plan and execute multi-step tasks independently — browsing the web, writing code, managing files, and delivering finished outputs. But a controversially opaque credit system, documented reliability issues, geopolitical instability following a blocked Meta acquisition, and a dismal G2 score make it a tool you should stress-test carefully before committing.
Core autonomous agent capability is real and praised by returning users, but a documented 25–35% overall failure rate, unpredictable credit consumption regardless of success/failure, and frequent 'high service load' errors severely undermine reliability and ROI accessibility for production use.
Trustpilot dominated by 1-star billing and refund complaints, a confirmed credit-drain bug where support refused refunds despite agent self-diagnosing malfunction, no SOC 2 or security certifications found, and an unresolved Meta acquisition blocked by Chinese regulators with co-founders under exit bans creates critical trust deficits.
Strong adoption signals with millions of users and $100M+ ARR, $75M Series B from Benchmark and Tencent at $500M valuation, and active recent integrations (Slack, Gmail, Notion, WhatsApp), but the unresolved China regulatory block of the Meta acquisition and ownership uncertainty dampen enterprise confidence.
API v2 is versioned and documented with webhooks, multi-turn conversations, and file management; community SDKs active in Go, PHP, TypeScript, and Python with recent 2026 commits; however, no official first-party SDKs confirmed, no OpenAPI spec found, and platform durability concerns exist given regulatory/ownership instability.
Manus arrived in March 2025 with a genuinely compelling pitch: an AI that doesn't just think but acts — planning multi-step tasks, browsing the web, writing and running code, and handing you a finished deliverable. The viral hype was real, the waitlist was enormous, and the benchmark numbers looked impressive. Then reality hit.
Frequently asked
What is Manus?
Manus is a genuinely impressive autonomous AI agent that can plan and execute multi-step tasks independently — browsing the web, writing code, managing files, and delivering finished outputs. But a controversially opaque credit system, documented reliability issues, geopolitical instability following a blocked Meta acquisition, and a dismal G2 score make it a tool you should stress-test carefully before committing.
How much does Manus cost?
Manus uses a paid pricing model. See the pricing breakdown above for current plans.
What is Manus best for?
Manus is best suited for productivity.
What is Manus's StackScore?
Manus scores 52/100 on Instawhat.ai's independent StackScore™ — a mixed pick — review the breakdown before committing.