conscious-sapphire
conscious-sapphire2y ago

Setting Up Tiered Subscription Models: Limiting Usage or Prompts

Hey, I hope someone can help. I'm looking to offer my chatbots for sale through a monthly subscription model, featuring three levels: Starter, Advanced, and Pro. Could anyone guide me on setting this up in Voiceflow? Specifically, is there a method to restrict token usage within a certain timeframe, or to cap the number of prompts submitted? Thanks 👊
7 Replies
Tico | Voiceflow Assistant
To set up a tiered subscription model for your chatbots in Voiceflow, you can leverage the token system to limit usage based on the subscription level. Here's a brief overview to guide you: - Voiceflow uses tokens as a currency for interactions with AI models. Each subscription plan offers a specific monthly token allocation. - Starter Plan includes 100,000 tokens, Pro Plan offers 2,000,000 tokens, and Teams Plan comes with 10,000,000 tokens. The Enterprise Plan provides a customized token quota. - If users exceed their monthly token quota, they can purchase additional tokens at $5 per 1 million tokens. This option is available for Pro, Teams, or Enterprise plans. - Tokens are consumed during interactions with AI models, with the amount depending on the features used and the complexity of the interactions. Advanced AI models like GPT-4 consume more tokens than earlier versions such as GPT-3.5. - To manage and monitor token usage effectively, consider the volume of users and interactions your chatbots will handle and choose the appropriate subscription plan to match your needs. For more detailed information on managing tokens and setting up agents, you can visit our guides on creating an assistant and understanding tokens in Voiceflow.
conscious-sapphire
conscious-sapphireOP2y ago
I need to limit the token usage further into 3 different tiers for the user using the bot. Is there a way to achieve this?
broad-brown
broad-brown2y ago
I would recommend using the usage API to check what consumend how much in what timeframe. https://developer.voiceflow.com/reference/querypubliccontroller_queryusage
conscious-sapphire
conscious-sapphireOP2y ago
Thanks a lot. i'll take a look at this
provincial-silver
provincial-silver2y ago
I also want this feature. It’s no different than the feature they use to limit our own usage as customers of voiceflow. We just need a way that enables us to limit usage for our customers
W. Williams (SFT)
You could actually use the Query API and you get the token data. You could add up the tokens and that would be pretty simple to limit a user's usage. You then could store the data in a DB or something like airtable.
optimistic-gold
optimistic-gold2y ago
@Matt Humphreys could you not setup a counter variable and place it after the prompt block and then a condition that says when prompts reach X, a user gets a message to notify them that they have reached their tier prompt quota? As for the tokens, you can limit those in the AI blocks anyway can’t you?

Did you find this page helpful?