Manus releases 1.5 with a rebuilt agent engine, two tiers, and one-prompt full-stack app building. Tasks run about 4x faster than spring figures, quality scores rise 15 percent, and reported satisfaction ticks up 6 percent, according to internal benchmarks.
The update targets builders who want research, coding, deployment, and analysis in one workflow. Manus can scaffold backends, set up authentication, attach databases, and ship AI features, then launch a live preview, probe like a user, and self-fix issues it detects. Visual editing lets you point to UI sections and describe changes. Built-in analytics, versioning, permissions, custom domains, and event notifications come standard.
Introducing Manus 1.5
— ManusAI (@manusai) October 16, 2025
Faster, better quality results. Unlimited context.
Build full-stack web apps, with real AI features, backends, user logins, custom domains and analytics.
Run your biz, start your side hustle, or just have fun.
If you can dream it, Manus can build it. pic.twitter.com/SWCghfMDoV
Availability splits by tier. Manus-1.5-Lite opens to all users today. Manus-1.5 is for subscribers, who also get a 50 percent discount on Lite credit spend. The iOS listing mirrors the jump to app-grade builds and cites the speed and context bumps under version 1.50.0.
What changed versus earlier builds: a re-architected runtime that can allocate extra reasoning and compute on harder tasks, plus an expanded single-task context so long jobs keep coherence. The new Library centralizes outputs, while Collaboration brings teammates into shared sessions to co-drive the agent.
The company positions Manus as a general agent that “turns thoughts into actions,” operating across web, mobile, and its hosted builder. The public site frames 1.5 as “from prompt to web apps,” highlighting end-to-end generation with login, database, and backend from a single description.
