Understanding Chat Credits on OnVerb

One of the key features of OnVerb is the ability to access and utilise multiple advanced AI language models like GPT-3.5, Claude, PaLM 2, and others. However, each of these powerful models has different computational costs associated with generating responses. To provide a simple and flexible way of using any model within OnVerb, we’ve implemented a universal chat credit system.

The way that AI services charge is based on the input we send to the AI, and the response that we receive back from the AI.

What are Chat Credits?

Chat credits are OnVerb’s universal currency that allows you to pay for AI-generated responses across all the language models available on our platform. Instead of dealing with different pricing structures for each model, chat credits make it easy to access any model’s capabilities with a single payment method, and one pool of credits.

The cost of each message depends on the amount of data (context) that you send with each message to the AI. The longer your system prompt, the more credits that it will cost per message. As your chat gets longer, the context increases and so chats become increasingly expensive.

The cost of input tokens (the context we send to the AI, including our new message), and output tokens (the response we get back from the AI) vary, so we split the cost of a message into two parts. The context (input), and the response (output)

How Do Chat Credits Work?

Chat credits are deducted from your account balance every time you receive an AI-generated response, with the amount deducted depending on the specific model used and the length of the response. More powerful models like GPT 4 that require significant computational resources, are more expensive and so they will consume more credits per response compared to models like GPT-3.5.

This system ensures you only pay for what you use while having the flexibility to choose the best model for your needs. Whether you need GPT-3.5’s versatility, Claude’s advanced reasoning, or PaLM 2’s multi-task capabilities, your chat credits make it easy to access them all.

Credit Cost per AI Model

The table below shows the current chat token costs per model.

AI ModelInput Cost
1 Credit =
Response Cost
Each response =
Message cost with
1500 input tokens
GPT 3.5 Turbo (16k context)16000 tokens1 Credit2 Credits (1+1)
Claude Instant (100k context)15000 tokens1 Credit2 Credits (1+1)
PaLM 2 Chat Bison (32k context)5000 tokens1 Credit2 Credits (1+1)
Claude 2.0 (100k context)1500 tokens8 Credit9 Credits (1+8)
Claude 2.1 (200k context)1500 tokens8 Credit9 Credits (1+8)
Mistral Medium (32k context)4000 tokens3 Credit4 Credits (1+3)
GPT 4 Turbo (128k context)1000 tokens10 Credit12 Credits (2+10)
Claude 3 Sonnet (200k Context)1500 tokens6 Credit7 Credits (1+6)
Tokens roughly equate to 3/4 of a word. 1000 words is generally 750 tokens. Read more

When you start a chat, and as the chat progresses, the cost is broken down to make it as clear as possible.

screenshot of the cost calculations on Onverb

Managing Your Chat Credits

You can purchase chat credits through your OnVerb account using a variety of payment methods. Credits never expire, so you can stock up when there’s a sale or special offer. The OnVerb dashboard provides clear visibility into your remaining credits and usage history so you can monitor costs.

For higher-volume usage needs, we also offer discounted subscription plans that provide a set number of credits each month along with other premium features. These subscriptions automatically renew your credits so you never run out mid-project. You get free access to Claude Instant and GPT 3.5 Turbo on our paid monthly plans

Universal Access, Simplified Pricing

With chat credits, OnVerb ensures you have flexible, affordable access to the latest AI breakthroughs as they become available. You can experiment with new models, compare their performance, and select the best fit for each task, all without worrying about complex pricing structures.

Whether you’re an individual user or a business, OnVerb’s chat credits make premium AI more accessible and easier to budget for. Try them out today and experience the power of customised AI assistants!