Pricing

Text Models (per 1M tokens)

Model
Input
Output
gpt-5-mini
$0.425
$3.40
gpt-5-nano
$0.085
$0.68
gpt-4.1-mini
$0.680
$2.72
gpt-4.1-nano
$0.170
$0.68
gpt-4o-mini
$0.255
$1.02

Feature Multipliers

Some features may increase costs by applying a multiplier to the base token pricing:RAG 1.0×, STM 1.05×, LTM 1.10×, Personalization 1.25×.

RAG: Retrieval-Augmented Generation, STM: Short-term Memory, LTM: Long-term Memory. Personalization uses profile/context analysis.

How credits work

  • Buy any amount of credits you want — $1 to $100+.
  • Credits convert to tokens. Your conversations use tokens.
  • Different voices/models use tokens at different rates (see pricing above).
  • No subscriptions. No auto-renew. Add more credits whenever you need.
  • You'll always see an estimated tokens preview before you pay.

Example: $10 in Credits (gpt-5-nano, max personalisation)

If you always used gpt-5-nano with the highest personalisation level (1.25×), $10 would give you about 10,456 complete message–reply pairs (about 10 million tokens total). This assumes each exchange is around 300 tokens in your message and 700 tokens in the reply. Shorter messages give you even more turns; longer ones give fewer.

What happens when credits run out?

Your chat will pause and we'll show a simple "Add credits" prompt—no surprises.

Rates may vary by conversation length/model.