AI chatbot billing and prepaid balance

Understanding AI Chatbot Billing in Deskwoot

Deskwoot's AI bot is designed to give you full control over your costs. Rather than charging you per message or per conversation, AI bot usage is billed based on tokens consumed through your own API key. This means Deskwoot itself does not add a per-message fee. You pay only for what your AI provider charges based on actual token consumption. This approach keeps pricing transparent and directly tied to how much your bot is used.

How Token-Based Billing Works

Tokens are the basic units that AI language models use to process text. Every message your bot receives and every response it generates consumes a certain number of tokens. Here's what affects your token usage:

• Message length: Longer customer questions and longer bot responses use more tokens.
• Conversation context: If your bot sends prior conversation history with each request to maintain context, this increases the token count per interaction.
• Model selection: Different AI models have different per-token pricing. More advanced models typically cost more per token.

Because you connect your own API key, all token charges are billed directly by your AI provider according to their published rates.

Monitoring Usage

Keeping an eye on your AI costs is straightforward. To track your AI usage:

1. Navigate to Settings > Bots in your Deskwoot dashboard.
2. Review the available metrics, including message counts, token usage, and estimated costs.
3. Check this dashboard regularly to spot unexpected spikes or trends in consumption.

Best Practices for Managing Costs

• Keep bot responses concise: Instruct your AI bot to provide clear, focused answers. Shorter responses consume fewer tokens and often deliver a better customer experience.
• Set spending alerts with your API provider: Most AI providers allow you to configure usage limits or billing alerts. This helps prevent unexpected charges.
• Review usage periodically: Check your token consumption in Settings > Bots at least weekly, especially after launching a new bot or updating its instructions.
• Optimize conversation context: Avoid sending unnecessarily long conversation histories with each request to reduce token consumption per interaction.

Troubleshooting Common Issues

• Bot stops responding: If your AI bot suddenly stops generating replies, check that your API key is still valid and that your prepaid balance or billing account with your AI provider has sufficient funds.
• Higher-than-expected costs: Review your token usage in Settings > Bots. A sudden spike may indicate unusually long conversations, an increase in traffic, or overly verbose bot instructions.
• Estimated costs seem inaccurate: The estimated costs shown in Deskwoot are approximations. Always cross-reference with your AI provider's billing dashboard for exact charges.

Frequently Asked Questions

Does Deskwoot charge me for each AI bot message?

No. Deskwoot does not charge per message. Your costs come from the tokens consumed through your own API key, billed directly by your AI provider.

What happens if my API key runs out of credits?

If your prepaid balance with your AI provider is depleted or your API key becomes invalid, the AI bot will not be able to generate responses. Ensure your account with your provider stays funded to avoid service interruptions.

Can I control how much the AI bot spends?

Yes. You can manage costs by writing concise bot instructions, monitoring usage in Settings > Bots, and configuring spending limits directly with your AI provider.