Is Jompter free like Spokenly?

Yes, on your Mac. Dictation, prompts and the MCP hand-off are free and unlimited. Screenshots, recordings, boards and Remote have free limits, and a one-time license removes them. Jompter's free tier needs a free account; Spokenly's site says its local dictation needs none.

Does Jompter run on Windows or iPhone?

Jompter runs on Apple Silicon Macs with macOS 13 Ventura or later. The iPhone app is a remote for your Mac, not a dictation app. If you need dictation on Windows, Linux or iPhone, Spokenly covers those.

How do I switch from Spokenly to Jompter?

Download Jompter and sign in with a free account, pick Local and download a speech model once, then hold Right ⌘ to dictate. You can keep Spokenly on its own hotkey for other apps. Turn on MCP to connect Claude Code, Cursor or Codex.

Yes, on your Mac. Dictation, prompts, the MCP hand-off, transforms, translation and Speak are free and unlimited. The heavier features are free within limits — 20 screenshots a day, 5 screen recordings, 3 boards and one 10-minute Remote session a day — and a $99 one-time license removes every limit.

Jompter

A Spokenly alternative that sends your screen with your words.

Jompter is a Spokenly alternative for when words are not enough: the screenshot, the annotation and the screen recording travel in the same prompt, and your agent reads all of it over MCP. Spokenly is free local dictation for Mac, Windows, Linux and iPhone, with an MCP server that lets Claude Code, Codex and Cursor ask for voice input.

Free on your Mac · no card · Apple Silicon, macOS Ventura+

The short answerJompter vs Spokenly

Platforms

JompterApple Silicon Mac

SpokenlyMac, Windows, Linux, iPhone

Transcription

JompterOn-device, works offline

SpokenlyOn-device or cloud

Output

JompterA full prompt for your agent

SpokenlyText, or voice input over MCP

Full comparison, 10 rows

Quick answer

Pick Spokenly if you want free local dictation on Mac, Windows, Linux and iPhone, with an MCP server that lets agents ask you for voice input. Pick Jompter if you want your agent to receive the whole picture: your words plus screenshots and key frames from a screen recording.

What Spokenly does well

Spokenly offers free local dictation with Whisper and Parakeet, cloud models if you want them, and apps for Mac, Windows, Linux and iPhone. Its MCP server lets Claude Code, Codex and Cursor ask you for voice input and speak answers back, which is a nice idea.

Where Jompter is different

Both have an MCP server; what the agent gets is different. Jompter hands the agent a finished prompt: your transcript, annotated screenshots and recording frames, with a task list it keeps updated. Spokenly's site doesn't mention screenshots or recordings, so for visual bugs you'd paste those separately.

A recording of the real Jompter app on a Mac — Recorded in the real Jompter app.

Same task, both apps

Tell Claude Code about a button that jumps when the page loads.

With Spokenly

1Take a screenshot with Shift+Command+4.
2Dictate the description into Claude Code with Spokenly.
3Switch back, paste the screenshot with Ctrl+V, and point at it in words.
4For a bug that moves, record the screen separately and describe what happened, because the agent can't watch the video.

With Jompter

1Hold Right ⌘ and describe the bug.
2Press ⌘⇧S mid sentence and box the button. The screenshot lands between your words.
3For a bug that moves, record it with Video Prompt. The key frames and your narration join the same prompt.
4Paste one line into Claude Code, and it reads the whole prompt over MCP.

How to switch from Spokenly

Download Jompter and sign in with a free account.
Pick Local in setup and download a speech model once: Whisper, Parakeet or Nemotron.
Hold Right ⌘ to dictate, or rebind it in Settings. Keep Spokenly on its own hotkey if you still want it for other apps.
Turn on MCP and connect Claude Code, Cursor or Codex, so your agent reads prompts itself.

Dictation gives you words. Coding agents need context.

Prompts, not just text.

Dictation tools stop at words. Jompter bundles your voice, screenshots, annotations, and screen recordings into one AI-ready Markdown prompt for Claude Code, Cursor, or ChatGPT.

Free on your Mac.

Dictation, prompts and the MCP hand-off are free and unlimited on your Mac, with no meter and no card. Screenshots, recordings, boards and Remote are free within limits, and a one-time license removes them, no subscription required.

Context that persists.

Plans keep the running context of a debugging session (the error, the screenshots, what you tried), so a follow-up is one line added, not everything explained again.

Speak the bug. Snap the screen. Paste one prompt.

Your words, the annotated screenshot and the key frames from a recording land in one Markdown prompt, ready for Claude Code, Cursor or ChatGPT.

Voice, on-device. Whisper, Parakeet and Nemotron transcribe on Apple Silicon, no audio leaves your Mac.
Screens, redacted. Annotate with pen, arrow, box, text or blur before anything is shared.
Copied in one click. Markdown and images together, straight into your agent.

Jompter vs Spokenly

Both tools are good at what they're built for. The question is what you're building prompts for.

CapabilityJompterSpokenly

Open sourceNot stated on its site

PriceFree on your Mac; $99 once lifts the limitsFree local dictation; paid Pro plan

PlatformsmacOS (Apple Silicon)macOS, Windows, Linux, iPhone

Speech models on your machineWhisper, Parakeet, NemotronWhisper, Parakeet (plus cloud models)

Voice input for Claude Code, Codex and Cursor over MCP

Screenshots + annotation inside the promptNot advertised

Screen recording → key frames + transcriptNot advertised

Persistent prompt workspaces (Plans)Not advertised

Agent reads saved prompts and to-dos over MCPNot advertised

Runs on Windows, Linux and iPhoneMac only

Based on each product's public positioning, checked September 2026. Spot something out of date? Tell us and we'll fix it.

Which one should you pick?

Different jobs, different tools. Here's the fair split.

Pick Jompter if…

Your bug reports need a screenshot or a screen recording, not just words
You want the agent to read a whole prompt, text, images and frames, over MCP
You work on an Apple Silicon Mac and want prompts kept in Plans

Pick Spokenly if…

You need dictation on Windows, Linux or iPhone as well as Mac
You only need your voice turned into text, with no visual context
You want local dictation without signing in to anything

Free on your Mac. $99 once to lift the limits.

Dictation on local models is free and unlimited. Cloud is optional and separate.

Dictation

Forever

$0no card

Local models · no word cap

License

Best value

—one-time

or — / year · on-device, offline

Cloud

Optional

—/ month

Heavy — / month · cloud transcription + AI transforms

Compare every plan on the pricing page →

Switching from Spokenly?

The questions developers ask before trying Jompter.

What the agent receives. Spokenly's MCP server lets an agent ask for voice input and speak answers back. Jompter's MCP server hands the agent a finished prompt: your words plus screenshots, annotations and key frames from a screen recording, with a to-do list the agent keeps updated.

Other tools developers switch from.

Jompter vs MacWhisper

A MacWhisper alternative that writes prompts, not transcripts.

Read the comparison

Jompter vs Wispr Flow

A Wispr Flow alternative built for AI coding. Free on your Mac.