AI tool comparison
Ghost Pepper vs VoxCPM2
Which one should you ship with? Here is the side-by-side panel verdict, pricing read, reviewer split, and community vote comparison.
Voice & Dictation
Ghost Pepper
Hold Control. Speak. Release. It types for you — all on-device.
75%
Panel ship
—
Community
Free
Entry
Ghost Pepper is a macOS hold-to-talk dictation app that runs entirely on-device using Apple's WhisperKit for speech recognition and LLM.swift for smart cleanup. You hold the Control key to record, release to transcribe, and the transcribed text is automatically pasted into whatever app you're using. No cloud, no subscription, no data ever leaves your Mac. The "smart cleanup" feature is what sets it apart from basic Whisper wrappers: it uses a local language model to remove filler words, fix self-corrections in real time, and clean up stutters without altering your intent. Version 2.0.1, released April 6, brings improved accuracy and lower latency on Apple Silicon. It requires macOS 14+ and an Apple Silicon chip. Ghost Pepper hit the top of Hacker News' Show HN section on April 7 with 354 points and 164 comments — an unusually strong signal for a solo-dev open-source tool. The timing is notable: as commercial dictation tools like Wispr Flow move to paid-only models, Ghost Pepper offers a fully free, auditable alternative. It's MIT-licensed and available on GitHub.
Voice AI
VoxCPM2
Describe a voice in text, get studio-quality speech — no reference audio needed
75%
Panel ship
—
Community
Free
Entry
VoxCPM2 is a 2B-parameter text-to-speech system from OpenBMB — the team behind MiniCPM — built around a tokenizer-free, diffusion-autoregressive architecture. Most TTS systems convert text to discrete audio tokens first, then decode those tokens to waveform. VoxCPM2 skips the tokenization step entirely, operating in continuous latent space. The result is 48kHz output with smoother prosody and finer pitch control than token-based systems. The headline feature is "Voice Design": you describe a voice in natural language — "a confident male voice, mid-Atlantic accent, slightly gravelly, deliberate pacing" — and VoxCPM2 synthesizes a brand-new voice from that description without any reference audio sample. This is architecturally different from voice cloning (which requires samples) and voice selection (which picks from a catalog). It supports 30 languages with automatic detection, no language tags required. The model runs on consumer hardware (~8GB VRAM), integrates with the MiniCPM-4 language model backbone, and is released under Apache 2.0. For developers building multilingual voice products or researchers exploring generative voice control, VoxCPM2 represents a meaningful step beyond current open TTS leaders like F5-TTS and CosyVoice.
Reviewer scorecard
“This is the dictation tool I've been waiting for. On-device, zero latency once warmed up, MIT license, and the LLM cleanup actually works. I replaced Wispr Flow with this in under 5 minutes. The Control-hold UX is more ergonomic than I expected.”
“The tokenizer-free architecture is the right technical move — eliminating the quantization artifacts from discrete audio tokens is the main reason commercial TTS still sounds better than open source. The Voice Design feature alone is worth experimenting with for anyone building voice products. 8GB VRAM requirement is very reasonable.”
“Apple Silicon only and macOS 14+ means a significant portion of Mac users are locked out. The 'smart cleanup' LLM adds another model to memory — not ideal if you're already running other local models. Also, no GUI means non-technical users won't touch it.”
“48kHz is great on paper, but the diffusion-based approach likely trades inference speed for quality. No benchmarks are published against F5-TTS or Kokoro in the README, which is a red flag. Voice Design sounds novel but natural-language voice descriptions are inherently ambiguous — you'll get inconsistent results across generations.”
“Ghost Pepper is a preview of how computing will feel in 5 years: ambient voice input everywhere, zero latency, zero cloud dependency. The fact that a solo dev shipped this in Swift using WhisperKit and LLM.swift is a testament to how capable the Apple Neural Engine stack has become.”
“Voice Design as a primitive changes how voice AI gets built. Instead of recording actors, teams can describe and iterate on synthetic voices the way designers iterate on color palettes. When this technology matures, every product that uses voice will have a unique, consistent, describable brand voice — not a voice cloned from someone else.”
“I tried it during a writing session and the filler-word removal alone is worth it — my raw dictation comes out cleaner than when I type. The hold-to-talk model also means I'm never accidentally recording. Solid privacy story for journaling and creative work.”
“Finally a TTS tool where I can describe what I want instead of auditioning samples. For narration, podcasts, and video, being able to say 'warm, unhurried, slightly husky' and get a consistent voice is a workflow unlock. The 30-language automatic detection is huge for multilingual content creators — no more manually tagging each segment.”
Weekly AI Tool Verdicts
Get the next comparison in your inbox
New AI tools ship daily. We compare them before you waste an afternoon.