What is BoltAI?
BoltAI is a chatbot UI built natively for macOS that connects to over 300 AI models through a single desktop app. It works with major AI providers and runs local models too. You bring your own API keys and pay providers directly for what you use. The app opens instantly over any other app with a keyboard shortcut. It handles text, images, documents, PDFs, and screenshots without switching windows. BoltAI also runs on iPhone and iPad, with chats, agents, and projects synced across devices.
Features & Benefits
- Multi-model support: connect to and switch between 300+ AI models from providers including OpenAI, Anthropic, Google, Mistral, Azure, and Amazon Bedrock, plus local models via Ollama and LM Studio, all from one chatbot UI.
- Image generation: generate AI images directly in chat using supported vision-capable models, with native QuickLook support for previewing results.
- Vision analysis: attach images, photos, screenshots, or UI captures and get explanations, summaries, or structured answers from any vision-capable model.
- Document Q&A: attach PDFs, research papers, or reference files and get structured answers drawn from the document content.
- Code execution and agent skills: run code and use agent skills to analyze data, generate documents, and automate tasks without leaving the chatbot UI.
- Inline workflow: build repeatable AI workflows that run inside any app on your Mac, with control over the template, agent, post-processing, and output layout.
- Screenshot to answer: capture your screen with a keyboard shortcut and get an instant AI response about what’s on it.
- Web search: run fast web searches directly from the chatbot UI without switching to a browser.
- AI agents: build custom agents with their own system instructions, knowledge base, and tools for consistent, task-specific outputs.
- Multi-agent chat: run multiple agents in parallel and compare their responses side by side.
- CLI integrations: connect existing subscriptions for Codex CLI, Claude Code, Gemini CLI, and GitHub Copilot and use them directly inside BoltAI.
- Configuration profiles: switch instantly between saved configurations for different tasks like coding, writing, or research.
- Project folders: organize chats into projects with shared context, system prompts, and tool settings that every new chat in that project inherits automatically.
- Fork and branch chats: split a conversation at any point to explore alternative directions without losing the original thread.
- Prompt library: save and reuse prompts with a keyboard shortcut for faster, more consistent chatbot UI sessions.
- Attachment management: add files, documents, or reference material to any chat and manage what’s included in the context through the Inspector pane.
- Advanced LLM controls: fine-tune temperature, max tokens, top-p, top-k, frequency penalty, and presence penalty for precise control over model behavior.
- Voice input and dictation: speak prompts into any text field using AI-powered dictation that works with both cloud and local models.
- Read aloud: have BoltAI read responses out loud for a hands-free experience.
- Global keyboard shortcut: open the chatbot UI instantly over any app with a single keystroke.
- Full-text search: search all past chats locally with fast on-device results.
- Chat export: export any conversation to Markdown or PDF for sharing or archiving.
- MCP support: connect to local and remote MCP servers, enable individual tools per agent, and configure custom runtime environments.
- Cross-device sync: use BoltAI on Mac, iPhone, and iPad with chats, agents, and projects kept in sync via optional cloud sync.
- Local model support: run LLMs locally via Ollama or LM Studio for fast, private responses with no per-token costs.
- Local data storage: chats are stored on-device in a local database and never used to train AI models.
- API key encryption: encrypt stored API keys with a custom passphrase so only you can access them.
Real-World Applications
A software developer working across multiple projects can use BoltAI’s chatbot UI to keep sessions organized by project folder. Each folder holds its own system prompt, attached files, and agent settings. Switching from a debugging session to a documentation task takes one click. Forking a chat to test an alternative prompt keeps the original thread intact.
Writers and content teams can use the inline workflow feature to run chatbot UI prompts directly inside other Mac apps. A blogger drafting in their writing tool can trigger a rewrite or summarization command without opening a separate window. The prompt library handles recurring tasks like tone adjustments or title generation without retyping the same instructions each time.
Researchers and analysts may find the document Q&A and vision features useful for processing dense material quickly. Attaching a PDF and asking the chatbot UI specific questions about it surfaces answers in seconds. Screenshot analysis works the same way for charts, dashboards, or UI elements that are easier to capture than describe.
Professionals in privacy-sensitive fields can run BoltAI with local models only, keeping every chatbot UI interaction fully on-device. Encrypted API keys and local chat storage mean no dependency on third-party infrastructure for sensitive work.
