Moltbot (formerly Clawdbot) is a powerful open-source AI assistant that can run locally and interact with you via various messaging platforms (like Telegram, Discord).

This article demonstrates the simplest script-based installation method, teaching you how to integrate it with Telegram while using Google Gemini’s free quota to enhance the experience.

1. Environment Preparation

Before starting the installation, please ensure your server or local environment meets the following minimum requirements. Although the script handles some dependencies, pre-checking avoids many errors.

In the previous article, we discussed how to choose a suitable VPS for Clawdbot (now named Moltbot). Today, we will officially get our hands dirty and teach you step-by-step how to deploy this powerful AI assistant to your server and interact with it anytime, anywhere via Discord.

This article is compiled based on practical sharing from AppSail, covering the complete “one-stop” process from server purchase to security hardening.

What is Moltbot?

Moltbot (formerly Clawdbot) is an open-source AI assistant developed by Peter Steinberger. Unlike commercial products like Claude Code, Moltbot runs on hardware (or VPS) controlled by you, connecting browsers and messaging platforms (such as WhatsApp, Telegram, Discord, Lark) via WebSocket architecture.

As Clawdbot (now renamed Moltbot) continues to gain popularity, many developers and AI enthusiasts are starting to look for suitable deployment environments. Should you invest in high-performance local hardware or choose a flexible cloud VPS? This has become a choice everyone faces.

This article, referencing practical experience from tech blogger AppSail, outlines the best low-cost schemes for deploying AI assistants in 2026.

Local Hardware vs. Cloud VPS: Crunching the Numbers

For many geeks, having a Mac mini M4 as a home server is a dream configuration. However, when we take out the calculator for a detailed check, reality might be a bit “leaner”.

“Google’s being generous again (I advise you stop giving credits, give more Banana quota first)”

For the vast number of AI developers and loyal Google users, there is good news today. Google AI Pro members (including memberships claimed via student verification) can now claim $10 API credit monthly on Google Cloud Platform (GCP).

This $10 might not seem like much, but it’s absolutely sincere for individual developers to test the latest models or run some small Agents. More importantly, this credit can be used to call Google’s most powerful models currently available, including the highly anticipated Banana Pro (Nano Banana Pro) and Gemini 3 Pro.

With Moonshot AI officially releasing Kimi 2.5 and its onboard k0-math reasoning model, Chinese large models have seen a qualitative leap in logical reasoning and mathematical problem-solving capabilities. For developers and geeks, being able to run such a powerful model in a local environment means not only higher data privacy security but also lag-free offline calls.

Although Kimi 2.5 is a massive Mixture-of-Experts (MoE) model, thanks to Ollama’s ecosystem support, we can now easily deploy its quantized version on personal computers.

In today’s fiercely competitive Chinese LLM market, Moonshot AI has once again delivered a blockbuster result.

Recently, Kimi welcomed a comprehensive version iteration (widely known in the industry as Kimi 2.5). This update is not just a stack of parameters, but a qualitative leap in reasoning capability and multimodal understanding by introducing Reinforcement Learning at the underlying architecture level.

1. Core Breakthrough: k0-math Reasoning Model

The most eye-catching focus of this update is undoubtedly the release of the k0-math model. This is Kimi’s first reasoning model adopting the Reinforcement Learning technical route, aiming directly at OpenAI’s o1 series.

In the field of AI coding assistants, simple “conversational” capabilities are gradually becoming standard, while real competition is shifting towards “execution” and “extensibility”.

A recent tweet by renowned developer Manthan Gupta on X (Twitter) has attracted widespread attention in the community. He pointed out that the next major leap for AI Agents will no longer just be an increase in model parameters, but a modular ecosystem built around “Skills”.

From “General Chat” to “Specialized Skills”

As Manthan Gupta implied, current AI coding assistants (like GitHub Copilot, Claude Code, Cursor), while powerful, are often limited by preset capabilities. Every time a user starts a new task, they often need to repeatedly input a large amount of context and prompts.

In the open-source community and AI field, the AI assistant project Clawdbot, built by renowned developer Peter Steinberger (founder of PSPDFKit), has been receiving much attention recently. However, just today, the project not only underwent a major rebranding but also unexpectedly got involved in a typical cryptocurrency scam.

Trademark Conflict: From Clawdbot to Moltbot

The story starts with a trademark issue. Although the name Clawdbot was loved by the community, due to its pronunciation being too similar to AI giant Anthropic’s flagship model Claude (Clawd vs. Claude), there was a potential risk of trademark infringement.

After the open-source AI assistant Clawdbot announced its rebranding to Moltbot and sparked heated discussions in the community, the reaction speed of domestic cloud vendors was truly surprising.

Within just a few days, the Tencent Cloud developer community quickly followed up, launching a nanny-level deployment tutorial for Moltbot. It even allowed many users to discover that by using Tencent Cloud’s Lighthouse, one can own a 24/7 online private AI butler for a cost of just ¥38/month.

In this era of AI explosion, owning a standby private AI assistant is no longer a plot from science fiction movies. Clawdbot is such a powerful open-source AI assistant that can run on your own device (Mac, Windows, Linux) and seamlessly integrate with the chat apps you use daily (such as Telegram, WhatsApp, Discord, etc.).

Best of all, you can access the Qwen (Tongyi Qianwen) model to get top-tier AI experience at extremely low or even zero cost.