Units of Intelligence (UoI)
ToothFairyAI uses Units of Intelligence (UoI) as a simple, transparent way to measure and pay for AI operations. Instead of complex per-feature pricing, UoI gives you a single currency to budget and track all your AI usage.
What is a UoI?
1 UoI = $0.01 USD
That's it. A clear, predictable unit that makes it easy to understand exactly what you're spending on AI operations. Whether you're generating text, creating images, or transcribing audio, everything is measured in UoI.
- Predictable costs — Know exactly what each operation costs before you run it
- Flexible allocation — Use your budget across any AI service
- No seat fees — You are billed for Units of Intelligence (UoI blocks), not per user
- Real-time tracking — Monitor usage as it happens
UoI and pre-paid credits
ToothFairyAI is a pure PaaS product — there is no fixed monthly subscription and no per-seat pricing. Instead, you buy intelligence credits upfront and you are billed for the Units of Intelligence (UoI) you actually consume. Credits are debited as your agents run, and the credits you pre-purchase never expire — you only ever pay for what you use.
Billed UoI = the UoI you actually consume, drawn down from your pre-paid credit balance
Because there is no headcount or seat count, unlimited users keep working under the same credit balance; adding people never adds a per-seat fee.
Tiers and how you pay
| Subscription | How you pay | UoI rate |
|---|---|---|
| Serverless | Pay as you go — buy credits upfront, billed for UoI consumed | 1 UoI = $0.01 USD |
| Enterprise | Custom annual commitment, with reserved UoI and reserved compute | 1 UoI = $0.01 USD (custom allocations available) |
- Serverless is consumption-based: top up your credit balance at any time, and your credits never expire. There is no fixed monthly minimum or maximum — you simply consume UoI as you go.
- Enterprise includes a custom UoI commitment and reserved compute, arranged with your account team. Enterprise UoI can be increased above the minimum commitment by contacting support.
Pricing by Service
Text Generation (LLMs)
Large Language Model usage is measured in tokens. Pricing is published live from the backend single source of truth and varies by model tier based on capability and performance. The table below is generated live from the public /models_list endpoint and updates automatically whenever the backend tier pricing changes — so the website never needs manual edits when models or margins are adjusted.
Loading live UoI pricing...
Image Generation
Create images with AI using various models at different quality and capability levels:
| Model Tier | Cost per Image | Models |
|---|---|---|
| Small | 105 UoI | FLUX Schnell, Stable Diffusion XL |
| Medium | 150 UoI | FLUX Dev, Stable Diffusion 3, Sorcerer Canvas, Sorcerer Vision, Mystica Vision |
| Large | 225 UoI | FLUX Pro, FLUX 1.1 Pro, Seedream 4, Ideogram V3 Turbo, Qwen Image |
| Extra Large | 300 UoI | FLUX 1.1 Pro Ultra |
Video Generation
Generate and analyse videos with AI:
| Model Tier | Cost per Video | Example Models |
|---|---|---|
| Small | 200 UoI | Fast generation models |
| Medium | 400 UoI | Standard quality models |
| Large | 700 UoI | High quality models |
| Extra Large | 1,000 UoI | Premium models (Veo 3, Kling) |
Voice Synthesis (Text-to-Speech)
Convert text to natural-sounding speech:
| Voice Model | Cost per 1M Characters |
|---|---|
| Kokoro | 1,200 UoI |
| Orpheus 3B | 4,500 UoI |
1 million characters is approximately 200,000 words or 400 pages of text.
Audio Transcription (Speech-to-Text)
Transcribe audio and video files to text:
| Transcription Model | Cost per Minute |
|---|---|
| Whisper Large v3 Turbo | 1.2 UoI |
| Whisper with Speaker Diarisation | 3 UoI |
3D Model Generation
Generate 3D models from text or images:
| Model Tier | Cost per Model |
|---|---|
| Medium | 150 UoI |
Monitoring Your Usage
Track your UoI consumption in real-time from your workspace:
- Dashboard — View current usage and remaining budget
- Settings > Notifications — Set alerts at 80% and 100% thresholds
- Usage History — Review detailed breakdowns by service type
When your credit balance runs out, requests will be paused until you top up your credits. Credits never expire, so any unused balance remains available in the next cycle — there is no monthly reset.
Best Practices
Optimise Token Usage
- Use concise prompts where possible
- Choose appropriate model tiers for each task
- Leverage smaller models for simple tasks
Plan Your Budget
- Start with smaller models and scale up as needed
- Monitor usage patterns to predict monthly needs
- Set up notifications to avoid unexpected limits
Choose the Right Model
- Simple queries → Small tier models (cost-effective)
- Complex reasoning → Large tier models (better results)
- Production workloads → Medium tier (balanced)
Need More UoI?
If you are running low on credits:
- Top up your credits — Buy more intelligence credits at any time. Credits never expire, so unused balance carries over indefinitely.
- Contact sales — Enterprise customers can request custom UoI allocations and reserved compute.
Topping up your credits
Serverless is pay-as-you-go: you buy intelligence credits upfront and are billed only for the UoI you consume. Credits never expire.
How to top up
- Open the Workspaces screen and edit your workspace
- Go to Billing or Credits
- Choose the credit package you want to purchase
- Confirm the purchase — your new credit balance is available immediately
- Serverless is billed purely for the UoI you consume from your pre-paid credit balance. Credits never expire — you only ever pay for what you use.
- Enterprise plans include a custom UoI commitment and reserved compute. Enterprise UoI can be increased above the minimum commitment by contacting support.
- Reserved compute workspaces cannot adjust UoI in-app — contact us to make changes.
For questions about UoI or optimising your usage, contact our support team or visit your workspace dashboard.