Compare
Presspeech vs Superwhisper
Both can transcribe on-device on Apple Silicon. Superwhisper is a configurable dictation workspace: model choice, AI post-processing modes, and Windows/iOS support, with free and paid tiers. Presspeech is a free single-purpose push-to-talk pipeline tuned for one thing — latency.
| Question | Presspeech | Superwhisper |
|---|---|---|
| Primary workflow | Hold hotkey, speak, release — the verbatim transcript pastes at the cursor about 100 ms later. | Record with a hotkey, then optional AI modes reshape the output (notes, messages, custom prompts) before insertion. |
| Engine path | Local Parakeet TDT v3 CoreML model on the Apple Neural Engine, decoded in a single pass on key release. | Choice of local Whisper-family models, local Parakeet models on the Pro tier, plus optional cloud models and LLM post-processing from multiple providers. |
| Privacy posture | No cloud path exists: no account, no telemetry, transcript history in memory only. | Local models run fully offline; the optional cloud models and cloud AI modes send data out when you choose them. |
| Platforms | Apple Silicon Macs, macOS 14 or later. | macOS, Windows, and iOS; Superwhisper notes Intel Macs work best with its cloud models. |
| Price | Free and MIT licensed. | Free tier with smaller local models; Pro is $8.49/month, $84.99/year, or a $249.99 lifetime license. |
| Configurability | Deliberately minimal: hotkey, trigger style, language pinning, local text corrections. | Extensive: modes, prompts, model management, vocabularies, and per-app behavior. |
Choose Superwhisper if you want model choice, AI-shaped output, or Windows and iOS coverage, and the tiers fit your budget. Choose Presspeech if you want a free tool that does exactly one job with the lowest possible friction and no cloud path at all.
Sources: Presspeech source, Presspeech benchmarks, Superwhisper official site and its published pricing (checked June 2026).