Code Dictator vs Wispr Flow vs Claude Code vs Copilot CLI: Free Voice Dictation for AI Coding in VS Code

Voice input for AI coding assistants got crowded fast in 2026. If you're picking a tool, here's a
direct, feature-level comparison — including where the tools we didn't build genuinely win.

Quick answer

Want free voice dictation that works with whatever AI assistant you're already using, right inside VS
Code? That's Code Dictator's whole pitch: free, no paid tier, and it pastes into Claude, Copilot, Codex,
Cline, or anything else with a text box. It's not the only reasonable choice, though — Wispr Flow's free
tier is genuinely usable if you want dictation across your whole OS, not just your editor, and VoiceTypr
is worth a look if you want to pay once and never touch a cloud API. Claude Code's /voice and Copilot
CLI's voice input are fine if you already live inside those specific CLIs and don't need dictation
anywhere else.

The options, side by side

Code Dictator

A VS Code extension: press Alt+D, talk, press Alt+D again, and your words land on the clipboard — ready
to paste into Claude, Copilot, Codex, Cline, or any other AI chat or terminal. Speech-to-text runs
through ElevenLabs (90+ languages), with a free-tier API key covering typical daily use. The extension
itself carries no license fee, ever — it's funded by an ElevenLabs affiliate referral and an optional Buy
Me a Coffee tip, not by charging users. It's been actively maintained since its March 2026 launch (15
point releases and counting).

The honest limitation: it's VS Code-only, and it depends on a cloud speech-to-text API (ElevenLabs by
default), so it's not the tool for people who want zero network calls.

Worth a fair mention: VS Code itself now ships its own built-in dictation too (Cmd/Ctrl+I in the chat
panel) — free, on-device, no cloud API call at all. The catch: it only writes into VS Code's own chat
input, editor buffer, or terminal, not the system clipboard, so unlike Code Dictator it can't dictate into
a browser tab or any other app outside VS Code.

Wispr Flow

The best-funded and most polished option here — Wispr Flow just raised $81M and markets itself as a
"Voice OS," not just a dictation app. It's system-wide (Mac, Windows, iOS, Android — dictate into Slack,
email, or any app, not just your editor), and community reports confirm it works inside VS Code too. The
free tier is real: $0/month, 100+ languages, MCP integration into Claude and ChatGPT. Pro is $15/user/month
for unlimited dictation and their more advanced models; Enterprise is custom-priced with SOC2/HIPAA
readiness.

If you want one dictation tool for your entire machine, not just code, Wispr Flow is the stronger pick —
its free tier is not a bait-and-switch. The tradeoff is that it's a general-purpose app that happens to
work in VS Code, not a VS Code-native extension, and unlimited use requires the $15/mo Pro tier.

VoiceTypr

The strongest counter-argument to "free is better." VoiceTypr charges $69 once — lifetime, 2 devices —
and markets explicitly against subscriptions ("most tools charge $10-15/mo, forever, VoiceTypr is one
payment"). It runs Whisper or Parakeet models locally by default (cloud is optional), which matters if
you don't want your voice going to a third-party API at all, and it ships a CLI and Agent API for wiring
dictation into your own tooling. The catch: macOS 14+ and Windows only — no Linux build.

For privacy-conscious or subscription-fatigued developers on Mac or Windows, this is a genuinely better
fit than any free tool here, ElevenLabs-cloud-based or not.

Claude Code's built-in /voice

If you already work inside the Claude Code CLI, /voice needs no separate install — it's bundled with
your Claude.ai login. The catch is real: it requires a local microphone, so it doesn't work in remote,
web, or SSH sessions, needs WSLg under WSL, and streams audio to Anthropic's servers (cloud-only, no
local option). It's also disabled outright if your organization runs Claude Code via a direct Anthropic
API key, Bedrock, Google Cloud Agent Platform, or Microsoft Foundry, or under an org HIPAA-compliance
policy. And it's scoped to Claude Code specifically — it won't help you dictate into Copilot or anything
else.

GitHub Copilot CLI voice input

The same idea, narrower still: a voice-input feature of the Copilot CLI, bundled with Copilot CLI access,
usable only there. Worth flagging honestly: an open, unresolved community thread ("did speech to text get
removed?") raises a real question about whether this feature is currently stable — we haven't verified it
either way, so treat that as an open question, not a claim in either direction.

Price isn't the whole story

Free is a genuine Code Dictator advantage over VoiceTypr's $69 and Wispr Flow's $15/mo Pro tier — but
it's not a clean win by itself. Wispr Flow's free tier is real and usable, not a crippled trial. And if
you care more about never sending audio to a cloud API than about the price tag, VoiceTypr's one-time,
local-first model is arguably the better deal long-term, not the worse one — you're trading a
subscription for a platform limitation (no Linux), not for less capability.

Which one should you actually use?

• You want free, VS Code-native, and works with any AI assistant → Code Dictator.
• You want dictation across your whole OS, not just your editor, and don't mind $15/mo for unlimited use → Wispr Flow.
• You want to pay once, run offline, and you're on Mac or Windows → VoiceTypr.
• You already live in the Claude Code CLI, on a local machine, and only need dictation there → Claude Code's /voice.
• You're all-in on Copilot CLI → its voice input — but check it's still active before you depend on it.

FAQ

Is there a free alternative to Wispr Flow for VS Code?
Yes — Code Dictator is a free VS Code extension with no paid tier. Wispr Flow also has a genuinely usable
free tier, but its unlimited/advanced-model usage sits behind a $15/month Pro plan.

Does Claude Code have built-in voice dictation?
Yes, via the /voice command — but it needs a local microphone (no remote/web/SSH sessions), streams
audio to Anthropic's servers, and is disabled under some enterprise configurations (direct API key,
Bedrock, Google Cloud Agent Platform, Microsoft Foundry, or org HIPAA policy).

What's the most private, offline voice dictation option for coding?
VoiceTypr, which runs Whisper or Parakeet speech-to-text locally by default. It's macOS 14+ and Windows
only — there's no Linux build.

Can I use voice dictation with GitHub Copilot in VS Code?
Copilot CLI has its own voice-input feature, but it's scoped to the CLI, not VS Code, and its current
stability is an open question in the community. Code Dictator works as a separate VS Code extension and
pastes into whatever you're using, including Copilot Chat.

Marketplace: https://marketplace.visualstudio.com/items?itemName=irrationalways.code-dictator&utm_source=devto&utm_medium=article&utm_campaign=play4-comparison
GitHub: https://github.com/Dmitriusan/code-dictator?utm_source=devto&utm_medium=article&utm_campaign=play4-comparison