ChatBoost - AI Chat Client

Muggle Studio

In-app purchases

About this app

ChatBoost is a clean, efficient AI client built for speed and flexibility.
No account or login required — just add your own API key and start chatting.

It supports multiple providers and models, including OpenAI, Azure OpenAI, Anthropic Claude, Google Gemini, OpenRouter,
Mistral, DeepSeek, Qwen (Alibaba), Bedrock, Ollama, and more.
Switch models anytime and fine-tune settings like Temperature, Top‑p, and Max Tokens.

Now with AI Keyboard:
Use ChatBoost AI across apps with an intelligent keyboard experience. Draft, rewrite, translate, and refine text
instantly while typing — without switching between apps.

Highlights:

1. 🔑 No login required — bring your own API key
2. 🌐 Multi-provider & multi-model support
3. ⌨️ AI Keyboard for writing, rewriting, and translating in any app
4. 🤝 AI agent with tool calling and optional web search
5. 💬 Streaming responses
6. 🖼️ Image input for multimodal/vision tasks
7. 🎙️ Voice input (speech recognition)
8. 🔊 Read-aloud AI replies (TTS)
9. 📚 Built-in assistant library
10. ✏️ Custom assistants & prompts
11. 🔎 Conversation list & global search
12. 💾 Offline-first local chat history
13. 📝 Markdown rendering
14. 🎨 Light / Dark / System themes with dynamic color

We value your feedback.
If you run into issues or have ideas, contact us at: [e-mail]

Ratings and reviews

I've been using this to chat with local LLMs on my network running on KoboldCpp. I bought premium so I can iterate on system prompts. It was working great. But after an update today, it can't seem to communicate with my KoboldCpp server anymore. It says CLEARTEXT communication is not supported by network security policy. What changed? My KoboldCpp server is not using HTTPs, it's just hosted on local network.

Good but the Xai api only shows models up to Grok 4.1 with 4.2 and even 4.3 already released and available on other apps

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

What’s new

• 🧠 Long chats automatically compress context for steadier follow-up replies
• ⚙️ Provider settings support editing each model's Content Window and Max output tokens
• 🧹 Provider settings support removing built-in models
• 🖼️ Assistant replies display Markdown images inline, with mixed text-and-image layouts
• 🔎 Added Bring search service (no API key required)