AI Usage & Quotas
Understand AI token consumption, quota limits, and how to monitor your AI usage in Enafeedback.
Last updated on
All AI features in Enafeedback — Insights, Transcription, and Translation — consume from a shared AI quota measured in tokens. Your quota resets at the start of each billing cycle.
AI quota by plan
| Feature | Approximate token cost |
|---|---|
| AI Insight (small data set) | 2,000–4,000 tokens |
| AI Insight (large data set) | 4,000–8,000 tokens |
| Transcription (per minute of audio) | ~1,500 tokens |
| Translation (per 100 words) | ~200 tokens |
Contact your account manager to see the exact quota included in your current plan.
Monitoring usage
- Navigate to Global Settings → AI Controls (requires
platform:adminrole). - The Usage section shows:
- Tokens consumed in the current billing cycle
- Tokens remaining
- Usage breakdown by feature (Insights / Transcription / Translation)
- Daily usage chart for the past 30 days
Usage alerts
Configure an alert when your AI usage reaches a threshold:
- In AI Controls → Alerts, set a percentage threshold (e.g., 80%).
- Select team members to notify.
- When usage exceeds the threshold, an email alert is sent.
This helps avoid unexpected quota exhaustion mid-month.
What happens when quota is exhausted
When the AI token quota reaches 0:
- Insights — The Generate button is disabled. A banner explains the quota status.
- Transcription — New audio attachments are stored but not transcribed until quota resets.
- Translation — New feedback arrives untranslated. Existing translations remain visible.
No data is lost. When the quota resets (or if you upgrade your plan), queued transcription jobs process automatically.
Quota overage
Your plan determines whether overages are allowed:
- Hard limit — AI features are disabled when quota is exhausted. No additional charges.
- Soft limit with overage — AI features continue; overage is billed at a per-token rate at end of cycle.
Check your plan terms for overage policy.
Caching
To maximise quota efficiency, Enafeedback caches AI results:
- Insights — Cached for 24 hours per module + date range combination.
- Translations — Cached indefinitely per source text.
- Transcriptions — Stored permanently after first generation.
Cached results never consume additional quota.
Disabling AI features
To preserve quota, individual AI features can be disabled:
- Go to Global Settings → AI Controls.
- Toggle off the features you want to disable (Insights / Transcription / Translation).
Disabled features do not appear in the UI for any team member.