SKI

Free voice coding for Claude Code, Codex and more

1.4K followers

Free voice coding for Claude Code, Codex and more

1.4K followers

Visit website

AI Dictation Apps

We just built SKI — voice coding for Claude Code, Codex and more. It's not dictation: your agent answers you out loud, like a real teammate, so you build at the speed you think. You can even bring it into a meeting to build live, or send it in your place to speak for you. It's an ambient thing that just sits on your desktop — hit a key, talk, it works. All on your machine, free. Available on Mac & Windows.

The Best SKI Alternatives

The best SKI alternatives are Wispr Flow, superwhisper, Granola, MacWhisper, and Szept.

Wispr Flow

4.8 ·

Choose Wispr Flow if...

✓you want voice typing inside any app
✓you dictate in mixed languages without toggling
✓you need desktop plus mobile dictation

See details ↓

superwhisper

4.9 ·

Choose superwhisper if...

✓you need offline dictation for sensitive work
✓you want custom modes for coding outputs
✓you prefer a lifetime purchase over subscriptions

See details ↓

Granola

4.8 ·

Choose Granola if...

✓your day is nonstop meetings and follow-ups
✓you want recording without a bot joining
✓you need notes mixed with transcript context

See details ↓

MacWhisper

4.9 ·

Choose MacWhisper if...

✓you mainly transcribe recordings and video files
✓you need fast subtitle generation workflows
✓you want reliable on-device transcription

See details ↓

Szept

Choose Szept if...

✓you want instant dictation with tiny app footprint
✓you use Claude Code and paste screenshots fast
✓you are fine using cloud STT via API key

See details ↓

What to Consider

SKI stands out for two-way voice conversation with coding agents, plus local meeting capture—more like a spoken teammate than a simple dictation button. But the alternatives split into distinct camps: “dictate everywhere” layers like Wispr Flow that drop formatted text into any app, offline-first power tools like superwhisper with programmable modes, meeting-native note systems like Granola that record without a bot joining, transcription workhorses like MacWhisper for turning audio/video into text and subtitles, and lightweight utilities like Szept that optimize fast agent handoffs (even screenshots) via cloud STT.

In comparing options, we weighed privacy and offline capability versus cloud convenience, latency and accuracy (including mixed-language/code-switching), how well each tool works across apps and devices, workflow fit for developers versus meeting-heavy teams, customization depth, resource/battery impact, platform coverage, reliability, and pricing models (subscription vs one-time).

Wispr Flow

Voice productivity for your writing and your meetings.

4.8 · 81 reviews

Learn more →

Wispr Flow is built for one thing SKI doesn’t optimize for: frictionless voice-to-text that drops into whatever app is already open. Instead of a spoken back-and-forth with coding agents, it behaves like an “everywhere” layer for emails, docs, tickets, prompts, and IDE comments—ideal when the goal is simply replacing typing at speed.

A major reason to pick Wispr Flow is mixed-language dictation that keeps embedded English product names and technical terms intact without constant language switching. That code-switching strength is especially useful for developers, PMs, and international teams who naturally blend languages while working.

Wispr Flow also wins when cross-device reach matters, since it’s used on desktop and mobile, including an iOS keyboard workflow that helped it earn Orbit Awards recognition in the AI dictation space. The trade-off versus SKI is that it’s not trying to be a voice agent teammate, and cloud processing and all-day background usage can be a concern for sensitive data or battery-heavy days.

Best for

Best for people who want fast dictation across all apps and devices, not a conversational coding agent.

Standout features

✓Dictation works inside most apps
✓Strong multilingual code-switching
✓Mobile support with iOS keyboard
✓Low-latency formatted text output
✓Cleanup and tone matching options

superwhisper

Extremely accurate, voice to text for Mac & iPhone

4.9 · 22 reviews

Learn more →

Privacy-first workflows are where superwhisper separates itself from SKI and most cloud dictation tools. It focuses on on-device, offline speech-to-text so sensitive work can stay local, while still behaving like a practical input method across everyday apps.

What makes it particularly compelling is the “modes” concept: users can define different dictation behaviors for writing, coding, or structured outputs, effectively turning dictation into a programmable pipeline. For developer workflows, that means a coding mode that shapes what gets produced before it ever hits the editor, rather than relying on a spoken agent conversation to refine it.

Superwhisper also caters to accessibility and power-use details, like flexible activation options beyond hotkeys and vocabulary customization for domain terms. If subscription fatigue is a factor, its lifetime purchase option can be a decisive advantage over tools that require ongoing plans.

The main trade-off is polish on mobile compared to desktop-first use, but for Mac users who want local control and deep customization, it’s one of the most capable alternatives to SKI.

Best for

Ideal for Mac users who want offline dictation with highly customizable modes for writing and coding.

Standout features

✓Offline on-device dictation
✓Custom modes and prompt rules
✓Vocabulary and terminology customization
✓Flexible activation options
✓Lifetime purchase option

Granola

The AI notepad for people in back-to-back meetings

4.8 · 56 reviews

Learn more →

Granola takes a meeting-native approach rather than a voice-agent or dictation-first approach like SKI. It’s designed as a notepad that blends what gets said with what gets typed, so the most important moments and your manual notes stay connected in one timeline.

A standout reason to choose Granola is the social dynamic: it records without a meeting bot joining, which keeps client calls and founder chats feeling normal and avoids changing participant behavior. That bot-free capture is a practical differentiator when professionalism and rapport matter as much as the transcript.

Granola also emphasizes in-meeting and post-meeting retrieval—chatting with what’s been discussed, generating summaries, and building a searchable memory across calls so the value compounds over time. Compared to SKI, it’s less about coding-side execution and more about turning conversations into durable decisions, action items, and recall.

The trade-offs are that speaker attribution and deeper semantic search can be less dependable than teams might want for accountability-heavy workflows, and heavy meeting days can raise CPU or battery considerations.

Best for

Best for meeting-heavy teams and operators who want bot-free recording with actionable notes and recall.

Standout features

✓No meeting bot joins calls
✓Typed notes interleaved with transcript
✓In-meeting chat for summaries
✓Cross-meeting search and recall
✓Actionable meeting summaries

MacWhisper

High Quality Text Transcription with OpenAI's Whisper on Mac

4.9 · 7 reviews

Learn more →

MacWhisper is the clear pick when the workflow starts with a recording, not a live conversation. Instead of optimizing for real-time voice interaction like SKI, it focuses on turning audio and video files into clean text quickly—great for meeting recordings, voice notes, podcasts, and content workflows.

It’s also a strong fit for on-device transcription fans who want a polished UI rather than wrangling command-line Whisper setups. That “transcription purist” positioning is exactly why MacWhisper earned Orbit Award recognition: it wins by being fast, straightforward, and dependable at the core job of transcription.

For creators, MacWhisper’s subtitle-friendly workflows are a major advantage, delivering fast transcripts suitable for captions and editing pipelines. Compared with SKI’s agent-centric experience, MacWhisper is simpler and more specialized—less about doing work in response to speech, more about producing an accurate transcript you can reuse anywhere.

The main limitation is that it’s not trying to dictate into every app or manage meetings end-to-end; it shines when batch transcription quality and speed matter most.

Best for

Ideal for creators and professionals who frequently transcribe recordings or generate subtitles on Mac.

Standout features

✓On-device transcription for audio and video
✓Subtitle and caption-friendly exports
✓High accuracy with large models
✓Simple, polished desktop UI
✓Fast turnaround for long recordings

Szept

Say it. Show it. Hand it to your agent.

Learn more →

Szept is built for developers who want an ultra-light dictation utility that stays out of the way. Where SKI leans into a full local, agent-driven voice workflow, Szept emphasizes speed and ergonomics—especially for working in terminals and agent environments like Claude Code.

Its signature advantage is context handoff: it can pair a quick screenshot with a spoken instruction and paste the right reference back into the originating terminal, helping an agent “see what you mean” without manual packaging. That makes it particularly handy for debugging, UI issues, and visual bugs where text alone is slow.

Szept also trades offline models for cloud streaming STT using the user’s own API key, keeping the app footprint tiny and latency extremely low. For users who prioritize responsiveness and minimal resource overhead over strict local processing, that’s a compelling counterpoint to SKI’s local-first philosophy.

The trade-off is clear: macOS-only and no offline mode, so it’s best when cloud transcription is acceptable and the goal is rapid agent prompting rather than a full voice conversation loop.