tara← Back

Models

Everything Tara and Tara Flow run on. On-device by default, always — the one cloud option below is off unless you turn it on yourself, and it only ever sends the single thing it needs to.

Step one

Speech-to-text

Whisper small.enOn-device
~465 MB
Tara — Converse & Dictate

English speech recognition, runs resident in memory for fast repeat transcriptions.

Parakeet TDT 0.6B v3On-device
~670 MB
Tara Flow

NVIDIA's model, int8 quantized. Tara Flow's own pipeline, independent of Whisper — downloads only if you turn Flow on.

Step two

Language model

Turns raw transcript into a clean reply or a tidied-up message — filler removed, punctuation fixed, self-corrections resolved. Four on-device options to choose from, plus one cloud option in Tara Flow.

Qwen3 4BOn-deviceDefault
~2.5 GB
Default — Converse, Dictate, Tara Flow tidy-up

Fastest replies, lowest memory use. No reasoning-mode toggle to fight — the 2507-Instruct line skips it entirely, which made it more reliable than the alternatives below at this size.

Qwen3 8BOn-device
~5.0 GB
Optional — swap in from Settings

Noticeably sharper than the 4B for about the same speed. Good upgrade if you're not sure which to pick.

Llama 3.1 8B InstructOn-device
~4.9 GB
Optional — swap in from Settings

Different training data and phrasing style than Qwen — worth trying if Qwen's voice doesn't feel right for you.

Phi-4 14BOn-device
~8.9 GB
Optional — swap in from Settings

Strongest reasoning of the four — best for tricky corrections or multi-step requests. Slower, more memory.

Gemini 3.1 ProCloudPaid · opt-in
Tara Flow — Smart Mode (optional)

Swaps Tara Flow's tidy-up pass to a larger cloud model for better quality. Your transcript leaves your Mac for this one request only — everything else about Tara stays exactly as private as always. Off by default; a paid feature when it's on.

Step three

Text-to-speech

KokoroOn-device
~325 MB
Tara — spoken replies

Open-weight text-to-speech, multiple voices to choose from in Settings.

Why this matters

Every model above except Gemini 3.1 Pro runs entirely on your Mac — nothing you say is ever sent anywhere by default. Smart Mode is the one deliberate exception, and it stays off until you turn it on. That's the whole cloud story, not a footnote to it.