Free
Try local/basic voice typing where supported
- Local baseline dictation
- Insertion into focused app
- Compatibility checker
- Per-app permission controls
Availability note: run the compatibility checker — plan does not guarantee platform support.
Start free where local/basic voice typing is supported. Upgrade when realtime recognition, translation, or supported command workflows are available for your setup.
We'll point at the smallest plan that covers it. Nothing locks in.
Compatibility availability note: run the checker before upgrading — a plan unlocks an entitlement, not a guarantee on your platform.
Try local/basic voice typing where supported
Availability note: run the compatibility checker — plan does not guarantee platform support.
Realtime recognition and translation where available
Availability note: run the compatibility checker — plan does not guarantee platform support.
Higher allowance and supported command workflows
Availability note: run the compatibility checker — plan does not guarantee platform support.
What each plan unlocks — independent of platform status.
| Feature | Free | Starter | Pro |
|---|---|---|---|
Local dictation & insertion Push-to-talk, insertion into focused app. | Verified | Verified | Verified |
Managed accuracy mode Opt-in higher accuracy via managed model. | Unsupported | Verified | Verified |
Realtime translation Verified language pairs only. | Unsupported | LimitedVerified pairs. | LimitedVerified pairs. |
Command mode Per-app permission, off by default. | Unsupported | Unsupported | GatedPer-app permission. |
Workflow templates | Unsupported | LimitedStarter pack. | Verified |
Example caps for the published v1 plans. Local dictation is always unlimited.
| Resource | Free | Starter | Pro |
|---|---|---|---|
| Managed dictation minutes / month | — | 300 | 1,500 |
| Translation requests / month | — | 500 | 3,000 |
| Command mode actions / month | — | — | Fair use |
| Local dictation | Unlimited | Unlimited | Unlimited |
Example caps shown for orientation. Final limits are confirmed at checkout; running over an allowance falls back to local baseline, it does not block dictation.
Paid tiers cover managed-model usage. The ladder reflects compute cost, not feature gating for its own sake.
Speech recognition runs on your CPU/GPU. We don't see the audio, we don't bill the compute. Dictation stays unlimited at every tier.
Higher-accuracy and translation models run on managed infrastructure. The Starter/Pro fee covers your share of that compute, with the allowance shown above.