DEV.co
MCP Servers · amicalhq

amical

Amical is an open-source, local-first AI dictation app built with TypeScript and Electron that converts speech to text using Whisper and formats output contextually based on the active application. It runs entirely offline with optional integration of Ollama for local LLM processing, eliminating cloud dependency and privacy concerns.

Source: GitHub — github.com/amicalhq/amical
1.4k
GitHub stars
127
Forks
TypeScript
Primary language
MIT
License (OSI-approved)

Key facts

Objective fields from the source. Values we can't verify are shown as “Unknown” rather than guessed.

FieldValue
Repositoryamicalhq/amical
Owneramicalhq
Primary languageTypeScript
LicenseMIT — OSI-approved
Stars1.4k
Forks127
Open issues46
Latest releasev1.9.3 (2026-06-30)
Last updated2026-07-07
Sourcehttps://github.com/amicalhq/amical

What amical is

Built on Whisper (speech-to-text), Ollama (local LLMs), Electron (desktop), Next.js (UI), and TailwindCSS, Amical provides real-time dictation with context-aware formatting. The stack includes Better-Auth for authentication, Zod for validation, and Turborepo for monorepo management, with Jest for testing and Fumadocs for documentation.

Quickstart

Get the amical source

Clone the repository and explore it locally.

terminalbash
git clone https://github.com/amicalhq/amical.gitcd amical# follow the project's README for install & configuration

Need it deployed, integrated, or customized instead? DEV.co ships production installs.

Best use cases

Privacy-Sensitive Dictation Workflows

Organizations or individuals requiring zero cloud transmission of audio/text data for compliance (HIPAA, GDPR, internal policy) can deploy Amical entirely on-premise with local Whisper and Ollama models.

Accessibility and Productivity Enhancement

Users with mobility constraints, RSI, or high-volume typing needs (developers, writers, researchers) gain 3x typing speed improvement with context-aware formatting that adapts to email, IDE, messaging apps without manual correction.

Offline-First Applications

Field teams, remote workers in low-connectivity regions, or airgapped environments benefit from fully functional dictation without relying on internet-dependent cloud transcription services.

Implementation considerations

  • Local model setup: Whisper and Ollama require one-time download of model weights (typically 1–10 GB depending on model size); plan disk space and initial setup time accordingly.
  • Hardware requirements: CPU/GPU specs for real-time transcription and LLM inference are not documented; test on target hardware (especially for M1/M2 Macs and GPU acceleration on Windows).
  • Hotkey conflicts: Custom hotkey configuration must be tested against existing system and application hotkeys to avoid capture/collision issues.
  • Data persistence and backup: Verify where notes/transcriptions are stored (local filesystem assumed) and implement user-owned backup strategy.
  • Electron app signing and distribution: Plan for code signing, auto-updates, and notarization (macOS) if deploying beyond individual use.

When to avoid it — and what to weigh

  • Requires Enterprise Support SLA — Project shows active development (latest release 2026-06-30, 46 open issues) but Unknown commercial support structure. Enterprise buyers needing guaranteed response times or dedicated support should Requires review vendor support terms.
  • Need Out-of-Box Multi-Language Support — Whisper supports multiple languages, but README does not detail which languages are pre-packaged or how non-English language models are provisioned. Verify language requirements before adoption.
  • Dependency on Proprietary Cloud Integration — If workflow requires seamless integration with cloud-based CRM, project management, or enterprise SaaS (Salesforce, Monday.com, etc.), current integrations appear limited to local MCP and Ollama; cloud sync/API integrations status Unknown.
  • Mobile-First or Cross-Platform Consistency — App is primarily desktop (macOS, Windows via Electron). iOS is in beta, Android available but README does not detail feature parity, sync strategy, or roadmap for mobile-to-desktop consistency.

License & commercial use

Licensed under MIT (MIT License), an OSI-approved permissive license. MIT grants rights to use, copy, modify, distribute, and sublicense the software with minimal restrictions; attribution required.

MIT license permits commercial use, modification, and distribution. However, commercial deployment (support, SLA, liability) is Unknown. If selling or building on Amical as a service, review any associated cloud infrastructure terms, warranty disclaimers, and consult legal counsel regarding liability for transcription accuracy or privacy incidents.

DEV.co evaluation signals

Editorial assessment — not user reviews. Directional, with an explicit confidence level.

SignalAssessment
MaintenanceActive
DocumentationAdequate
License clarityClear
Deployment complexityModerate
DEV.co fitGood
Assessment confidenceHigh
Security considerations

Project claims 'privacy-first: works offline' and uses local Whisper/Ollama, eliminating cloud transmission. Considerations: (1) Electron app surface: Electron framework has known attack vectors; verify latest Electron version is used and auto-updates are enabled. (2) Local model integrity: download Whisper/Ollama models from official sources to avoid tampered models. (3) Hotkey and accessibility APIs: app requires OS-level permissions (microphone, app focus detection) which could be abused; audit permission requests. (4) No mention of encrypted local storage or keychain integration for sensitive notes; data at rest may be plaintext on disk. (5) Third-party audit history Unknown. Do not assume security without independent assessment.

Alternatives to consider

OpenAI Whisper (CLI) + Python scripting

Free, open-source transcription engine. Requires manual setup and scripting; no UI, context-awareness, or hotkey automation. Better for developers comfortable with CLI; not a packaged app.

Dragon NaturallySpeaking (proprietary)

Industry-leading accuracy and deep OS integration (Windows primarily). Requires license and cloud components; closed-source. Best for enterprise users needing vendor support and advanced grammar/formatting.

Otter.ai (cloud-based SaaS)

Polished UI, real-time collaboration, cloud storage. Requires internet and sends audio to cloud; privacy trade-off. Suitable if team collaboration and cloud sync are priorities.

Software development agency

Build on amical with DEV.co software developers

Amical offers a compelling open-source alternative for privacy-sensitive dictation and transcription workflows. Engage with the maintainers via Discord to understand enterprise readiness, deployment options, and long-term roadmap before large-scale adoption.

Talk to DEV.co

Related open-source tools

Surfaced by semantic similarity across the DEV.co open-source index.

Related on DEV.co

Explore the category and the services that help you build with it.

amical FAQ

Does Amical work completely offline?
Yes, core dictation (Whisper) and optional LLM processing (Ollama) run locally. Initial setup requires downloading model weights, but once cached, the app requires no internet. Optional cloud integrations are planned but not yet released.
What languages does Amical support?
Whisper supports 99+ languages. README does not specify which are pre-packaged or optimized. Verify your language is available in Whisper's official model variants and test locally before production use.
Can I self-host or deploy Amical in a corporate environment?
MIT license permits self-hosting and deployment. However, corporate deployment (authentication, audit logging, multi-user sync, compliance integrations) is Unknown. Review architecture, audit requirements, and coordinate with vendor via Discord for enterprise-grade setup.
What is the cost?
Amical is free and open-source. There is no stated commercial license or premium tier. Local model downloads (Whisper, Ollama) are free but consume disk/bandwidth. Deploying to organization scale may incur infrastructure costs depending on setup.

Software developers & web developers for hire

DEV.co is a software development agency delivering custom software development services to companies building on open source. Our software developers and web developers design, integrate, and ship production systems — spanning web development, APIs, AI, data, and cloud. If amical is part of your mcp servers roadmap, our team can implement, customize, migrate, and maintain it.

Evaluate Amical for Your Organization

Amical offers a compelling open-source alternative for privacy-sensitive dictation and transcription workflows. Engage with the maintainers via Discord to understand enterprise readiness, deployment options, and long-term roadmap before large-scale adoption.