Price

Generation costs vary depending on the selected model and content type.

• Text models are billed per token.
• Image generation is priced per image.
• Text-to-speech is priced per 1,000 characters.
• Speech recognition is priced per minute of audio/video.
• Video generation is priced per second

Tokens are chunks of text, where 1,000 tokens equal approximately 750 English words. The entire context (all previous messages + the new message) is used to calculate context cost, and the latest generated message is used to calculate generation cost.

What is a token?

Tokens are chunks of text — 1,000 tokens equal approximately 750 English words. The entire context (all previous messages plus the new one) is used to calculate context cost, while only the latest generated message is used to calculate generation cost.

Character-based languages like Japanese or Chinese may use more tokens than expected.For example, “こんにちは、元気ですか？” has 9 characters but counts as ~14 tokens. This impacts generation costs across languages.

Moderation

What is content moderation?

Content moderation is the process of reviewing and filtering user-submitted information in AI systems. Its goal is to ensure the safe and ethical use of AI technologies.

Why is moderation necessary?

Some AI model providers, like OpenAI and Anthropic, require mandatory content moderation to:

• Prevent illegal or harmful content
• Protect users from inappropriate information
• Ensure ethical use of AI
• Maintain quality and safety in AI interactions

How does moderation work on our platform?

We automatically moderate all content sent to AI models that require mandatory checks.Our system analyzes user inputs and blocks inappropriate content.

If prohibited content is detected:

• An error message will appear on the website
• The API will return an appropriate error code
• Repeated violations may result in access being blocked for moderated models!

Important! In case of repeated violations or upon request from the AI model provider, we are required to permanently block access to moderated models without the possibility of reinstatement.

How can I tell which models require moderation?

For user convenience, all models requiring mandatory moderation are marked with a special “Moderation” flag in the pricing list and model descriptions.

Usage Guidelines

• Carefully review the usage rules for each model
• Do not submit illegal, offensive, or harmful content to AI systems
• If you receive a moderation error, review and adjust your request
• Use AI responsibly and ethically

Content moderation is a key tool for ensuring safe and productive use of artificial intelligence technologies. By following the rules, you help advance this promising field and get the most out of your AI experience.