Skip to main content
Credit use depends on the operation, the selected models, the number of prompts, and the countries, personas, and boost configured for each prompt.

Collection methods

Qwairy collects answers through two methods:
  • Core models read the public interface of an AI assistant, such as chatgpt.com. Alexa for Shopping costs 2 credits per collected answer. The other Core models cost 1 credit per collected answer. Availability depends on the country and on the provider’s current service status.
  • Premium models call a named model through the provider’s official API. They cost 2 to 40 credits per answer, depending on the model. Native web search or grounding is enabled where the provider supports it.
These methods observe different product surfaces. Keep the collection method and model selection stable when comparing periods.
Beta models are collected on a best-effort basis, and Qwairy may disable them at any time. Credits are only charged for collected answers. Unavailable models are excluded from generation.

Core models

Meta AI (Core) is a separate beta available to enabled workspaces in France, Germany, Italy, Spain, the United Kingdom and the United States. When available in Monitoring, it costs 1 credit per collected answer. Check your workspace before including it in an estimate. Country support differs by model. See Supported countries and territories.

Premium models

Premium API models are enabled on request, with a tailored offer.

Previous-generation models

With premium API access enabled, these models remain selectable in Workspace Settings > Monitoring at their previous credit cost. They are not offered during onboarding.

Other credit costs

Fact Check does not consume customer credits.

Calculate a run

For each prompt, one run costs:
A prompt without personas counts as one persona, and a prompt without a boost counts as one answer per model. The countries are the prompt’s selected countries, or the workspace default country when none is selected. For 10 prompts on ChatGPT (Core), GPT-6.1 Sol, and Perplexity (Core), in the default country only:
If the same 10 prompts also run in a second country where all three models are available:
A single prompt with a boost of 3 answers on the same three models costs:
For monthly estimates, Qwairy counts about 30 runs for a daily frequency, 4.33 runs for a weekly frequency, and 1 run for a monthly frequency. For example, 100 prompts monitored weekly on ChatGPT (Core), Gemini (Core), and Perplexity (Core) use about 1,300 credits per month. The Monitoring page shows an estimate based on the current prompt frequencies, models, countries, personas, and boosts. Review that estimate before saving a change.

Check and replenish the balance

Your remaining balance appears in the dashboard sidebar and under Team Management > Usage. Credits are added by plan allocations, purchases, and grants. If the balance covers only part of a scheduled run, Qwairy may process an affordable subset of prompts or models. At zero credits, scheduled monitoring is skipped and manual generation is unavailable. The team Owner can buy a one-time credit pack or change plan under Team Management > Billing. Packs require an active paid subscription. Unused credits carry over while the paid subscription stays active. If you cancel at the end of the billing period, you can use the balance until that date. When the subscription ends, the entire remaining balance expires, including credits bought in packs. Sub-team credits are allocated by the parent team.