Beyond Chat: 39 AI Agents That Actually Do the Work.
Introducing chatweb.ai
Former Mercari CPO Yuki Hamada launches A2A v1.0-compatible multi-agent AI platform with self-improving feedback loops. Enterprise plan (¥1M/mo, limited to 10 companies) now available.
The Problem with AI Chat
When you tell ChatGPT to "send an email," it doesn't actually send one. When you ask it to "check my calendar," it can't. chatweb.ai changes that. 39 specialized AI agents actually send emails via Gmail, operate web browsers, execute code, publish websites, and automate workflows — all from a single natural language instruction.
Demo
chatweb.ai interface — Chat, Features, Workflow, API Documentation
What is chatweb.ai?
chatweb.ai is a multi-agent AI platform developed by Enabler, Inc. (CEO: Yuki Hamada). Users give instructions in natural language, and semantic routing automatically selects the optimal agent from 39 specialists. Multiple agents collaborate autonomously — research → analysis → report → email delivery — completed in a single prompt.
How It's Different
| ChatGPT / Claude | chatweb.ai | |
|---|---|---|
| Text generation only | Actually sends via Gmail | |
| Web Research | Limited search | Playwright browser + screenshots |
| Code | Sandbox only | E2B execution + results |
| Site Publishing | Not possible | xxx.chatweb.ai instant deploy |
| Scheduling | Not possible | Cron automation |
| Multi-Agent | Single model | 39 agents collaborate |
| LINE / Telegram | Not possible | Mobile messenger access |
Key Features
12 AI Models
Choose from Claude Opus 4.6, Claude Sonnet, GPT-4o, Gemini 2.5 Pro, Llama 4 Scout, Kimi K2, and more. Cost per message displayed transparently. Default Llama 4 Scout costs ~$0.0002/message.
A2A Protocol v1.0
Fully compliant with Google's Agent-to-Agent protocol v1.0 (now under Linux Foundation). JSON-RPC 2.0 endpoint enables seamless interoperability with 150+ organizations in the A2A ecosystem — including ServiceNow, Salesforce, SAP, and Microsoft.
Use cases:
- External CRM agent asks "research this lead" → chatweb.ai's Research AI executes → returns results via A2A
- Internal chatbot asks "review this contract" → chatweb.ai's Legal AI analyzes clauses → returns risk report
- Workflow engine asks "visualize sales data" → chatweb.ai's Analyst AI generates charts → returns report
Open Source
chatweb.ai is open-source (AGPL-3.0). We believe everyone should be able to build and run their own AI agent team — with full control over security, integrations, and data.
Self-Improving Feedback Loop
Every 6 hours, AI automatically analyzes quality metrics. When an agent receives 3+ negative ratings, the LLM generates and applies improved prompts — persisted across deployments. The system literally gets better while you sleep.
Quality Assurance
Plan-based quality targets (Free: 7/10, Pro: 9/10) with up to 5 improvement loops per response. Temperature, quality target, and loop limits are user-configurable.
Pricing
| Plan | Price | Credits | Highlights |
|---|---|---|---|
| Free | $0 | $2/mo | 39 agents, 12 models, E2B code execution |
| Pro | ¥2,900/mo | $19/mo | Opus/Gemini Pro, quality 9/10, cron automation |
| Team | ¥9,800/mo | $65/mo | Team workspace, shared agents |
| Enterprise (10 companies) | ¥1,000,000/mo | Custom | Dedicated AI consultant, custom dev, SLA 4h |
Enterprise Plan (Limited to 10 Companies)
At ¥1,000,000/month, a dedicated AI consultant analyzes your business workflows, designs optimal agent configurations, and handles implementation. Includes custom agent development, weekly meetings, dedicated Slack/Teams support, 4-hour SLA, and 99.9% uptime guarantee.
Why We Built This
LLMs are smart enough now. MCP and API integrations have become easy. But giving a single agent every capability is genuinely hard. The most natural architecture — the one that made the most sense to me — was a team of specialists. Big tech giants are trying to do this, but nothing has shipped yet. So I built it myself. I can connect the services I actually want, verify the security with my own eyes, and iterate at my own pace. chatweb.ai is open-source (AGPL-3.0). I want to create a world where anyone can have their own AI agent team.
About the Founder
Yuki Hamada — Engineer, entrepreneur, and angel investor with 20+ years of product development experience.
- Mercari (2014-2021): Director, CPO/CINO — Led product growth for Japan's largest marketplace app
- NOT A HOTEL (2018-2024): Former External Director — Built tech-driven hospitality experiences
- Cybridge (2003-2013): Co-founder — Web development agency
- Reiwa Travel (Newt): External Director — Travel tech startup
- Currently: CEO of Enabler, Inc. Building 10+ products including chatweb.ai, StayFlow (1,860 users), and Elio (offline AI). Angel investor
Try chatweb.ai for free
Get Started →Company Information
| Company | Enabler, Inc. (株式会社イネブラ) |
| CEO | Yuki Hamada |
| Location | Tokyo, Japan |
| Founded | 2024 |
| URL | https://chatweb.ai |
Media Inquiries
Enabler, Inc.
Email: info@chatweb.ai
Web: https://chatweb.ai