BrowseWiz - your AI Chat, Assistant and Agent
Free
Best option to try it out.
- Contextual AI chat
- Chat history
- One-click prompt collection
- YouTube summaries
- Multi-file analysis
- Bring your own key or model
- Custom instructions
- Limited custom tools
- 50,000 credits daily*
- 10 web searches per day
Advanced (Early adopter!)
Best for most demanding users.
- Everything from Free plan
- Unlimited custom tools
- Email support for customization
- Unlimited credits monthly**
- Unlimited web searches**
* See credits usage table for details.
** May be rate limited.
What is the value of credit?
Sending messages
Large Language Model (LLM) services used in the BrowseWiz application measure text length in tokens. One token is usually around 4 characters (but in rare cases, it can be as low as 1 character).
The credit cost of sending a message to an LLM service depends on the quantity of input and output tokens, and it is calculated with the following formula:
The table below shows input and output token costs in credits for each LLM model.
| Model Name | Credits per Input Token | Credits per Output Token | Input Token Limit | Output Token Limit |
|---|---|---|---|---|
| gemini-2.0-flash | 1 | 4 | 1,048,576 | 8,192 |
| gemini-2.5-pro-exp-03-25 | 0 | 0 | 2,097,152 | 8,192 |
| gpt-4o-mini | 1.5 | 6 | 128,000 | 4,096 |
| claude-3.5-haiku | 8 | 40 | 200,000 | 8,192 |
| deepseek-r1-distill-qwen-32b | 2 | 3 | 64,000 | 8,000 |
Input tokens sent with message include chosen context (active page text, video transcript, included file or URL), additional instructions based on chat mode (assistant, summarizer, writer - no more than 500 tokens), message entered by user, as well as all the historical messages displayed in the chat.
Example calculation
You are watching a ~20min commentary video on YouTube website. The transcript has around 30,000 characters. Assuming average 4 characters per token, this means around 7,500 of input tokens.
Response text has around 2,000 characters, which is roughly 500 output tokens.
You are using gemini-1.5-flash model, so the calculation is as follows:
After using each query, you can check the number of tokens you used in the ⚙️ > "General settings" in your Application.
Before using the query, you can also check the estimated number of tokens of a text snippet at the link (specifically for OpenAI models).