Pricing & Credits Policy

This page explains how access, features, and usage are billed. Pricing will evolve; treat this as a living policy.

Final pricing is not yet formalized. This policy outlines the intended structure for planning and transparency.

Components of pricing

  • Fixed access fee: unlocks the app and base features
  • Tiered features: higher tiers unlock advanced capabilities
  • Usage-based credits: variable cost tied to AI processing and related usage

Credits system

Credits represent usage and are consumed per AI prompt based on token counts.

Benchmark model

  • A benchmark model (e.g., GPT-4o) provides the reference cost per input and output token.
  • Each supported model has a multiplier relative to the benchmark (provider/model dependent).

Credit calculation (conceptual)

  1. Measure tokens for input and output
  2. Compute cost at benchmark rates
  3. Apply model multiplier
  4. Deduct resulting credits from the user’s balance

input_tokens = 500 output_tokens = 300 benchmark_rate_in = X credits per input token benchmark_rate_out = Y credits per output token model_multiplier = M (e.g., 0.5 for cheaper, 2.0 for pricier)

credits = (500 * X + 300 * Y) * M

=== Notes
```text
- Rates (X,Y) and multipliers (M) are subject to change as providers update pricing.
- We may normalize small operations to a minimum charge to cover fixed overhead.


## Balances and limits
- Users maintain a credits balance visible in the navbar and account page
- Workflows enforce `aiConfig.max_credits_per_post` to bound per-item spend
- When a balance is low, the UI surfaces warnings; at zero, AI-dependent features stop

## Tiers (indicative; subject to change)
- Starter: base scraping, limited AI, lower concurrency
- Pro: expanded AI, higher concurrency, Node Editor access
- Enterprise: custom limits, SSO, priority support

## Refunds and adjustments
- Adjustments may occur when providers retroactively change pricing or if anomalies are detected
- Support can credit accounts for verified issues

## FAQ
- How do I estimate monthly cost? Track average tokens per prompt and multiply by frequency; apply your tier and model multiplier.
- What if I switch models mid-run? Each prompt is charged at the active model’s multiplier at execution time.