ChatBoost: AI Agent Client

Muggle Studio

In-app purchases

About this app

A more flexible AI client, built for your phone

Tired of switching between apps whenever you want to use a different AI model?

ChatBoost brings multi-model chat, AI agents, web search, tool calling, and an AI Keyboard into one mobile app. Connect the AI services you already use, choose the right model for each task, and keep your conversations, files, and everyday work in one place.

ChatBoost is an AI client, not a hosted model service. No ChatBoost account is required, but before chatting you must add and enable a provider, enter an API key or endpoint if required, and select a model for the default assistant. Chat history stays on your device, and requests go to the services you choose.

Use multiple AI models in one app

Different models are useful for different tasks.

Connect OpenAI, Claude, Gemini, OpenRouter, DeepSeek, Qwen, Mistral, Ollama, Azure OpenAI, Amazon Bedrock, NVIDIA, and custom OpenAI-compatible services.

Choose a model for writing, switch to one with web search for research, or use a vision-capable model for images and files. Your conversations no longer need to be scattered across different apps.

Build AI agents for the way you work

Start with a built-in agent for writing, translation, learning, or research. You can also create your own with custom instructions and prompts.

Choose a model for each agent, customize its name and avatar, and enable web search or tool calling when the selected model supports them.

Find current information on the web

Give supported models access to web search so they can find and organize up-to-date information.

Use ChatBoost to:

• Find and summarize recent information
• Compare information from different sources
• Research topics and create content
• Extract web pages and organize key points
• Include current information in answers

Some models also provide native search. You can configure other search services when needed.

Let agents take action

With tool calling, supported AI agents can carry out tasks such as creating calendar events, generating images, or running available tools.

ChatBoost also supports MCP for connecting additional external tools.

On Android, you can set ChatBoost as your system digital assistant for faster access.

Use AI Keyboard in other apps

ChatBoost AI Keyboard brings AI to wherever you are already typing:

• Write and reply to messages
• Rewrite awkward sentences
• Translate between languages
• Change tone and writing style
• Polish emails, notes, and social posts
• Draft replies using the context of a conversation

Use AI inside messaging, email, browser, and note-taking apps without repeatedly copying text.

Work with text, images, files, and voice

ChatBoost supports more than text chat.

Upload images for compatible vision models to understand, use supported models to generate images, and add documents, audio, or video as conversation attachments. You can also speak your prompts and listen to responses with text-to-speech.

Long text can become an attachment, while code and Markdown stay readable.

More control when you need it

For supported models, adjust settings such as Temperature, Top-p, Max Tokens, reasoning effort, and streaming responses.

You can also:

• Add custom models and API endpoints
• Manage multiple API keys
• Search, pin, and organize conversations
• Review token usage
• Import, export, and back up data
• Customize fonts, themes, and voice settings

Configure a provider and default model before chatting; advanced settings remain available when you need more control.

Your models, data, and setup stay under your control

ChatBoost does not require a product account and does not lock you into one model or provider.

API keys and chat history are managed on your device. Required content is sent to the model, search, or tool service you select and handled under that service's terms and privacy policy.

Questions or feedback? Contact us at [e-mail].

Ratings and reviews

The app works well. I was facing few issues earlier but both were quickly resolved by the developer. Kudos to them 😊 Tried premium but the web search tool doesn't seem to be working. I'm using deepseek v4 flash with thinking disabled. Also wanted to know if there's a way to view token usage for a particular chat, as in total input and output tokens. That would be useful.

‎ Great app. But wasted opportunity! Lacks a lot of features: 1. Gemini API support for: - "Code_Execution" (Python interpreter); - "Google_search_grounding" tool; - Url_context; All these tools should be available, as they're all API based itself. 2. Incompatible with reasoning models. - The thoughts are merged WITH main response itself! 3. Unable to upload files. (only imgs can be uploaded... why?). Updated Review: Latest release fixes a lot of stuff. Gemini API support still absent (1).

Good app but need to add minimax, xiaomimimo, glm for API provider.

What’s new

• 📋 Copy or share all content directly from the Select text page
• ✉️ Send messages or quick prompts while AI is replying
• ⚡ Anthropic now supports prompt caching
• 🔗 Improved search citation display
• 🛠️ Fixed known issues