The Rundown AI homepage

Independent tool overview

BoltAI at a glance

BoltAI is a native, local-first AI workspace for macOS that connects to hundreds of cloud and local models through users' own provider credentials, then adds projects, agents, document chat, system-wide shortcuts, MCP tools, workflows, and optional cross-device sync.

Visit the official BoltAI site ↗
BoltAI product preview
Product type
Native multi-model AI client and workspace
Primary platform
macOS 13+ on Apple Silicon or Intel
AI access
Bring your own API keys, supported CLI subscriptions, or local models
Model credits
Not included; cloud-provider usage is billed separately
Local models
Ollama and LM Studio supported
Starting license
$79 one-time for Essential
License model
Perpetual app license with one year of updates and eligible sync access

Overview

What BoltAI is

BoltAI is an AI client rather than a model provider. It gives Mac users one interface for services such as OpenAI, Anthropic, Google Gemini, Mistral, Azure OpenAI, Amazon Bedrock, xAI, and other compatible providers, plus local models through Ollama and LM Studio. Users bring their own API keys or supported CLI subscriptions, choose the model for each chat or agent, and pay model providers separately for usage.

The app goes beyond a menu-bar chat window with project folders, custom agents, multi-model comparison, document Q&A, screenshots, dictation, inline workflows, code execution, a prompt library, chat branching, and local or remote MCP servers. Its local-first design is useful for privacy-conscious workflows, but privacy still depends on configuration: prompts sent to a cloud model are governed by that provider, optional Cloud Sync stores selected data remotely, hosted features may pass through BoltAI infrastructure, and powerful tools can access data or perform actions within the permissions users grant.

Use cases

Who BoltAI is best for

The strongest fit depends on the job you need the product to complete, not the size of its feature list.

Mac power users

Reach multiple AI providers from a native app, global shortcut, screenshot workflow, or inline text field without changing browser tabs.

Developers and technical teams

Compare models, configure generation parameters, connect MCP tools, use CLI providers, and work with code or technical documents.

Researchers and analysts

Organize long-running projects, attach source documents, fork conversations, and use different models for retrieval, synthesis, and critique.

Privacy-conscious individual users

Keep chat history local by default and use local models for workflows that should not be sent to a cloud inference provider.

Multi-provider buyers

Use one interface across several model vendors instead of committing every task and workflow to a single AI subscription.

Capabilities

Core BoltAI features

1

Multi-model chat

Switches among hundreds of supported models from major cloud providers, compatible endpoints, CLI integrations, and local runtimes.

2

Bring-your-own-key access

Connects directly to configured model providers using the user's own API accounts and billing.

3

Local model support

Works with Ollama and LM Studio for on-device inference without per-token cloud charges.

4

Instant Chat and global shortcuts

Opens an AI overlay above other Mac apps for quick questions and context-light tasks.

5

Screenshots and multimodal inputs

Sends screenshots, images, PDFs, presentations, documents, and other supported attachments to capable models.

6

Dictation and inline workflows

Supports voice input and reusable transforms that can run against text in existing applications.

7

Projects, folders, and persistent context

Groups chats with inherited system instructions, documents, tools, knowledge, and model settings.

8

Custom and multi-agent workflows

Builds agents with specialized instructions, knowledge, models, and tools, and can compare several agents or models side by side.

9

MCP tools

Connects to local and remote Model Context Protocol servers and lets users enable individual tools and runtime settings.

10

Prompt library and profiles

Saves reusable prompts and switches between configuration profiles for different tasks or clients.

11

Advanced model controls

Exposes system prompts, context limits, temperature, output tokens, reasoning effort, top-p, top-k, and supported penalty settings.

12

Conversation management

Includes full-text search, chat forking, message steering, read-aloud, persistent attachments, and Markdown or PDF export.

13

Optional Cloud Sync

Syncs eligible chats, projects, agents, prompts, settings, and assets across licensed devices, with storage determined by tier.

14

Mobile companion

Eligible Pro, Pro+, and team licenses include limited mobile access while BoltAI's mobile apps remain in beta.

Process

How the BoltAI workflow works

  1. Step 1

    Choose the privacy boundary first

    Classify the work and decide whether it can use a cloud model, must use an approved enterprise endpoint, or should remain on a local model.

  2. Step 2

    Configure only approved providers

    Add a restricted provider key or supported CLI account, review that provider's retention and training settings, and avoid broad administrative credentials.

  3. Step 3

    Protect credentials

    Use Apple Keychain or BoltAI's passphrase-controlled encryption, do not paste secrets into chats, and rotate keys if a device or configuration is exposed.

  4. Step 4

    Create a bounded project

    Set a narrow purpose, approved instructions, minimal documents, a known model, and only the tools needed for that workflow.

  5. Step 5

    Test the workflow safely

    Start with non-sensitive examples, inspect exactly what context is sent, confirm output quality, and check tool behavior before using real work.

  6. Step 6

    Verify important outputs

    Check claims against source files, run code in an isolated environment, and require human approval for legal, financial, medical, security, or externally published work.

  7. Step 7

    Control tools and sync

    Enable MCP tools individually, use read-only permissions where possible, and turn on Cloud Sync only when its storage and retention model fits the data.

  8. Step 8

    Maintain and back up

    Export important work, keep an independent backup, update the app and local runtimes, review provider usage costs, and remove stale keys, tools, and synced data.

Cost

BoltAI pricing and free plan

BoltAI sells a perpetual license with one year of app updates and tier-eligible Cloud Sync and mobile access. Cloud-model API credits are not included. After the first year, the installed local app continues working; renewal for updates and sync is optional and is described as 60% of the original pre-discount purchase price.

Free trial

Free trial

A limited evaluation of the native app before buying a license.

  • No account required to download the trial
  • Basic chat and selected evaluation features
  • Advanced workflow, multimodal, agent, and tool features are limited

Essential

$79 one-time

A single-seat license for one desktop device.

  • All app features and a perpetual local license
  • 512MB Cloud Sync storage during the active Updates & Sync period
  • One year of updates included

Pro

$99 one-time

The listed bundle price for professionals using more than one desktop and a mobile device.

  • Two desktop activations and one mobile device
  • 1GB Cloud Sync storage
  • All app features and one year of updates

Pro+

$199 one-time

A higher-device and higher-storage license for power users.

  • Three desktop activations and two mobile devices
  • 5GB Cloud Sync storage
  • All app features and one year of updates

Team Perpetual

$99/seat, 5-seat minimum

A company license with shared storage and priority support.

  • One device per seat and at least five seats
  • 10GB shared team storage and beta mobile access
  • Optional renewal listed at $79 per seat per year

Pricing checked . Check current pricing at the source ↗

Assessment

BoltAI strengths and limitations

Where it stands out

  • Native Mac interface with system-wide shortcuts and low context-switching overhead
  • Broad provider choice prevents the workspace from being tied to one model vendor
  • Bring-your-own-key pricing lets users control provider accounts and select models by task
  • Local-first chat storage and local-model support offer a stronger privacy option than cloud-only clients
  • Projects, agents, tools, prompt profiles, and conversation branches support repeatable professional workflows
  • Direct access to model parameters is useful for users who want more control than consumer chat apps expose
  • MCP support makes the workspace extensible beyond chat and document Q&A
  • Perpetual local license avoids a mandatory recurring app subscription
  • Optional sync means users can keep an entirely local workflow when cross-device access is unnecessary
  • A 30-day money-back policy and free trial reduce the risk of evaluating the paid app

What to consider

  • The app license does not include AI tokens; API, hosted-model, search, image, transcription, and other provider charges are separate
  • BoltAI is primarily a Mac product; Windows is described as coming soon, and mobile access is still beta
  • A perpetual license covers the installed version, but continued updates, Cloud Sync, and eligible mobile access require optional renewal after year one
  • The currently listed Pro price is presented as a bundle price and may change, so buyers should verify checkout pricing
  • Local-first does not mean every workflow stays local: cloud-provider requests leave the device, and hosted or proxy features may use BoltAI infrastructure
  • Each connected AI provider applies its own privacy, retention, training, regional, availability, and acceptable-use terms
  • Enabling Cloud Sync can upload chats, agents, memories, settings, files, attachments, and some encrypted credential fields
  • Deleting cloud data does not automatically remove local copies, backups, model-provider records, or data held by other connected services
  • MCP servers and code-execution tools can read files, call services, or take actions according to their permissions; untrusted servers or broad access create material risk
  • Generated answers can be inaccurate, outdated, biased, or fabricated regardless of which interface sends the prompt
  • Local models vary widely in quality and require sufficient memory, storage, compatible runtimes, model files, and device performance
  • Multi-model workflows can increase cost and make it harder to understand which provider received which context
  • API-key management adds operational responsibility for budgets, rate limits, rotation, revocation, and breach response
  • Sensitive workplace use still requires employer approval, an acceptable provider contract, data classification, and compliance review
  • No AI client should autonomously approve high-stakes decisions or publish consequential outputs without qualified human review

Compare

BoltAI alternatives

The right alternative depends on the specific output, workflow, controls and budget your project requires.

Business Operations

Raycast

A broader Mac productivity launcher with integrated AI, extensions, commands, and system-wide shortcuts.

Explore Raycast

Business Operations

ChatGPT

A simpler first-party option for users who prefer bundled model access and do not need bring-your-own-key multi-provider control.

Explore ChatGPT

Project Management

Claude

A first-party Anthropic workspace for users who primarily want Claude models, projects, artifacts, and provider-managed access.

Explore Claude

Questions

BoltAI FAQs

What is BoltAI?

BoltAI is a native AI workspace for Mac that connects to cloud AI providers, compatible endpoints, supported CLI subscriptions, and local models, then adds projects, agents, document chat, shortcuts, workflows, and MCP tools.

Is BoltAI the same as Bolt.new?

No. BoltAI is a multi-model desktop AI client for Mac. Bolt.new is StackBlitz's browser-based AI website and app builder.

Does BoltAI include ChatGPT or Claude credits?

No. BoltAI is bring-your-own-key by default and does not bundle model credits. Cloud API usage is billed by the provider, though supported CLI integrations may use eligible provider subscriptions.

How much does BoltAI cost?

Current one-time prices start at $79 for Essential. The vendor lists Pro at a $99 bundle price, Pro+ at $199, and Team Perpetual at $99 per seat with a five-seat minimum. Taxes and provider usage are separate.

Is BoltAI a subscription?

The core app uses a perpetual license. The purchase includes one year of updates and eligible sync access. After that, local use of the installed version continues, while renewal is optional for new updates, Cloud Sync, and eligible mobile access.

Are BoltAI chats private?

Chats are local by default, and BoltAI says it does not use customer content for model training. However, prompts sent to cloud providers follow those providers' policies, and Cloud Sync or hosted features process additional data remotely.

Can BoltAI run models locally?

Yes. Official materials list Ollama and LM Studio support. Local-model speed and quality depend on the selected model, runtime, and Mac hardware.

What are the system requirements?

The current download supports macOS 13 or newer on Apple Silicon and Intel Macs. Windows support is listed as coming soon, and the mobile apps remain in beta.

Is it safe to connect MCP servers?

Only connect servers you trust, inspect their configuration, enable the minimum set of tools, restrict file and account permissions, and require confirmation for consequential actions. MCP access can be much more powerful than ordinary chat.

Bottom line

Our BoltAI verdict

BoltAI is a strong fit for Mac power users who actively compare models, manage their own provider accounts, and want agents, documents, shortcuts, tools, and local models in one native workspace. Its local-first design and perpetual license are meaningful advantages, but buyers must account for separate provider costs and must configure privacy deliberately: the safest setup uses minimal credentials, approved providers, tightly scoped tools, local inference for sensitive work, optional sync only when needed, and human review for consequential outputs.

Visit BoltAI website ↗
The Rundown University

AI training for the future of work.

Get access to all our AI courses, hundreds of real-world AI use cases, live expert-led workshops, an exclusive network of AI early adopters, and more.

AI Courses

Get unlimited access to all of our current & upcoming industry-specific AI courses for the duration of your subscription.

Daily Guides

To keep up with the rapid pace of AI, our team publishes AI implementation guides daily. Our library contains 300+ practical use cases to automate real-world work.

Workshops

Join weekly, live, interactive sessions with industry leaders who are at the forefront of AI for hands-on implementation guidance and exclusive insights.

Community

Network with an exclusive community of AI-first professionals who are working smarter with AI. Learn how early adopters are using AI in their work and businesses.