Token-based Credits

Text models across every PigletGo product are billed by actual usage. When you pick a specific model, each call's token usage is converted to Credits at the model's official price. There are no per-request, per-slide or multiplier charges.

Auto does not use your Credits. Every account gets a free Auto allowance each day and each week, which refills at the start of the next period; your dashboard shows what is left as a percentage. Image generation is billed per image and is not covered here.

Conversion rate

$1 of model usage = 30 Credits, so 1 Credit = $0.0333. Models priced in CNY are converted at ¥1 = $0.15 first.

How Credits are calculated

1. Split a call's tokens into three buckets: uncached input, cached input, and output.
2. Each bucket costs tokens ÷ 1,000,000 × that bucket's official price (USD per million tokens).
3. Add the three to get the call's USD cost, then multiply by 30.
4. Round up to the nearest 0.01 Credit. Any call with usage costs at least 0.01 Credit.

You are charged at the price of the model you selected, regardless of which route serves the call. Reasoning tokens count as output.

Example: DeepSeek V4 Flash ($0.3/M input, $1.2/M output) with 10,000 input and 1,000 output tokens: input $0.003 + output $0.0012 = $0.0042, × 30 = 0.13 Credits.

Model credit rates

The table below shows regular uncached input and output usage per 10,000 tokens, before rounding.

Cache pricing

Cached input tokens use 20% of the original input Credits. The cached portion is separated first, while uncached input continues to use the regular input column.