I had a pretty interesting weekend playing around with Manus AI.
I'll be going over this in detail tomorrow during the Accelerator call, but both of these are pretty great.
My main gripe is that they are currently pretty expensive, so I can't experiment as much as I want to.
I went through both my Manus and Genspark credits pretty fast haha.
I found some pretty cool Manus projects on X that I will be sharing.
From Perplexity:
Recent Developments (March 2025):
- Paid Subscriptions & Mobile App: Launched two premium tiers: $39/month (3,900 credits, 2 concurrent tasks) and $199/month (19,900 credits, 5 tasks, priority access). An iOS app was also released
- Technical Upgrades: Backend AI framework upgraded to "toropic's 37 Sonnet," though early tests indicate slower response times compared to rivals like DeepSeek and ChatGPT
- Enterprise Focus: Operates as an invite-only beta, prioritizing enterprise clients for tasks like resume screening and stock analysis. Critics warn this exclusivity may hinder mass adoption
- Growth: Discord community surged to 170,000+ members post-launch, fueled by viral demos showcasing its ability to build custom websites and handle multi-step workflows
Genspark
April 2025 Launch – Super Agent:
- Voice Call Capability: Differentiates itself with AI-powered phone tasks, such as booking restaurants using a synthetic voice. This feature targets everyday users and international audiences.
- Architecture: Employs a "Mixture-of-Agents" system combining 8 LLMs, 80+ tools, and proprietary datasets to automate workflows (e.g., turning YouTube videos into slideshows).
- Funding & Valuation: Raised $100M in Series A (February 2025), reaching a $530M valuation. Previous $60M seed round in June 2024 valued it at $260M.
- Search Engine Innovations: Uses GPT-4o, Claude 3.5 Sonnet, and Gemini 2.0 Flash to generate "Sparkpages"—single-page summaries for complex queries. A "Deep Research" mode analyzes 1,000+ sources per query.
Competitive Landscape
- Manus vs. Genspark: While Manus excels at technical tasks (e.g., data analysis), Genspark emphasizes accessibility with voice interactions and broader tool integration.
- Challenges: Both face skepticism about real-world reliability. Genspark’s opaque agent system and Manus’s slower performance are under scrutiny.
- Market Trajectory: Manus plans to open-source parts of its system in late 2025, while Genspark aims to expand its reasoning-optimized LLMs.