Pricing Guide and AI Credit Usage
- July 5, 2026
- 7 mins read
Table of Content
Understand how Wize AI credits function, how token usage is calculated, and how businesses can estimate the costs of customer interactions and knowledge base training.
Additionally, Wize AI Agent uses AI credits whenever it processes customer messages, generates replies, retrieves knowledge, or trains data from your website, documents, FAQs, policies, and other business content.
Understanding AI credits helps you plan your monthly usage, estimate costs, and choose the right credit package for your business.
What Is an AI Credit?
An AI credit is the usage unit used by Wize AI. Instead of charging separately for every token, message, or knowledge base action, REVE Chat converts AI usage into credits.
For Wize AI :
| Usage Item | Token / Credit Value | Bangladesh Cost | USA Cost | India Cost |
|---|---|---|---|---|
| 1 AI Credit | 20,000 tokens | ৳1.20 | $0.01 | ₹1 |
| 1 AI Message Reply | Around 10,000 tokens | ৳0.60 | $0.005 | ₹0.50 |
| 2 AI Message Replies | Around 20,000 tokens | ৳1.20 | $0.01 | ₹1 |
This means if the AI Agent replies once to a customer, it typically uses around 0.5 AI credit. If a conversation includes multiple AI replies, the total credit usage increases based on the number of messages, length of the conversation, and complexity of the response.
What Is a Token?
A token is a small unit of text that the AI model reads and processes. A token can be a word, part of a word, punctuation mark, number, or space depending on how the sentence is written.
For example, a short customer question such as:
“Do you deliver outside Dhaka?”
uses fewer tokens than a long message such as:
“I want to order three products from your website, but I need to know whether you deliver outside Dhaka, how long it will take, and whether cash on delivery is available.”
The second message uses more tokens because it contains more words and more information for the AI Agent to understand.
Tokens are used when the AI Agent:
- Reads the customer’s message
- Understands the intent
- Searches the knowledge base
- Follows the instruction or workflow
- Generates a reply
- Uses previous conversation context
- Performs an action through connected tools or automation
AI Credit Pricing by Market
The cost of AI credits may vary depending on the market or billing region.
| Usage Item | Bangladesh | USA / Global USD Market | India |
|---|---|---|---|
| 1 AI Credit | ৳1.20 | $0.01 | ₹1 |
| 100 AI Credits | ৳120 | $1 | ₹100 |
| 500 AI Credits | ৳600 | $5 | ₹500 |
| 1,000 AI Credits | ৳1,200 | $10 | ₹1,000 |
| 5,000 AI Credits | ৳6,000 | $50 | ₹5,000 |
| 10,000 AI Credits | ৳12,000 | $100 | ₹10,000 |
Conversation Credit Estimate
A customer conversation usually includes multiple messages between the customer and the AI Agent. The more times the AI Agent replies, the more credits are used.
For estimation purposes:
1 AI reply = around 0.5 AI credits
So, if the AI Agent replies 4 times in a conversation, the estimated usage is:
4 replies × 0.5 credit = 2 AI credits
Conversation Credit Estimate Table
| Conversation Type | Typical Length | Average Credit Usage |
|---|---|---|
| Very Short | Question answered in 1–2 AI replies | 0.5 – 1 credit |
| Quick Support Chat | Question answered in 3–5 AI replies | 1.5 – 2.5 credits |
| Normal Customer Conversation | Customer asks follow-up questions, 5–8 AI replies | 2.5 – 4 credits |
| Detailed Assistance | Product, order, pricing, or troubleshooting discussion, 8–12 AI replies | 4 – 6 credits |
| Long Sales / Complex Support | Multiple questions, qualification, comparison, troubleshooting, or handoff, 12–20 AI replies | 6 – 10 credits |
🔶 Note: Longer conversations naturally use more credits. Credit usage can also increase if the AI Agent needs to read more context, search a large knowledge base, use workflow actions, or provide detailed answers.
What Can Increase Conversation Credit Usage?
Not every conversation uses the same amount of credits. A short FAQ answer may use very few credits, while a complex sales or support conversation may use more.
Credit usage may increase when:
- The customer sends long messages
- The AI Agent gives detailed answers
- The conversation has many follow-up questions
- The AI Agent checks multiple knowledge base sources
- The AI Agent uses workflow automation or connected tools
- The AI Agent needs previous conversation history to understand the context
- The response includes product suggestions, policy explanation, order information, or troubleshooting steps
For example, a simple question like “What are your business hours?” may only need one AI reply. But a question like “Can you compare these three plans and suggest which one is best for my team?” may require a longer answer and more token usage.
Knowledge Base Training Credit Estimate
Wize AI Agent can be trained using website pages, documents, policies, FAQs, product information, and other business content. Training allows the AI Agent to understand your business and answer customer questions more accurately.
When you add content to the knowledge base, Wize AI processes that content and converts it into AI-readable knowledge. This training process also uses tokens and AI credits.
For estimation:
| Training Volume | Estimated AI Credit Usage | Cost in USD | Cost in BDT | Cost in INR |
|---|---|---|---|---|
| 1,000 words | 0.142 AI token | $0.00142 | ৳0.17 | ₹0.142 |
This means training 1,000 words of content uses a very small amount of AI credit. However, if you train an entire website with thousands of pages, the total usage will increase based on the total word count.
Website Training Estimate AI Token
The following table gives an estimated idea of how many credits may be required to train a website based on total word count.
| Estimated Words- Website Size | AI Token Used | Cost in USD | Cost in BDT | Cost in INR |
|---|---|---|---|---|
| Small Website- 100,000 words | 14.2 AI tokens | $0.142 | ৳17.04 | ₹14.20 |
| Medium Website- 250,000 words | 35.5 AI tokens | $0.355 | ৳42.60 | ₹35.50 |
| Large Website- 750,000 words | 106.5 AI tokens | $1.065 | ৳127.80 | ₹106.50 |
🔶 Note: These are estimated training costs only. Actual usage may vary depending on page structure, duplicate content, formatting, hidden text, tables, product descriptions, metadata, and how much content is included during crawling.
Common Page Training Estimate AI Token
Many businesses do not need to train an entire website at the beginning. You can start by training important pages such as FAQs, privacy policy, terms and conditions, product information, return policy, shipping policy, and support articles.
| Page Type- Estimated Words | AI Token Used | Cost in USD | Cost in BDT | Cost in INR |
|---|---|---|---|---|
| Privacy Policy- 2,000 words | 0.284 AI token | $0.00284 | ৳0.34 | ₹0.284 |
| Terms & Conditions- 3,000 words | 0.426 AI token | $0.00426 | ৳0.51 | ₹0.426 |
| FAQs- 1,500 words | 0.213 AI token | $0.00213 | ৳0.26 | ₹0.213 |
Recommended Knowledge Base Training Approach
For best results, start with the most important customer-facing content first. This helps your AI Agent answer common questions accurately without training unnecessary pages.
Recommended first training sources:
- FAQs
- Pricing information
- Product or service pages
- Shipping and delivery policy
- Return and refund policy
- Terms and conditions
- Privacy policy
- Support documentation
- Troubleshooting guides
- Contact and business information
After the first training, you can gradually add more content based on customer questions and AI performance reports.
Real Usage Examples
The following examples show how businesses can estimate monthly AI credit usage based on conversation volume.
These are estimates only. Actual usage may vary depending on message length, response detail, conversation complexity, and AI Agent configuration.
Example 1: Online Ecommerce Store
An ecommerce store uses Wize AI Agent to answer product, delivery, return, and order-related questions.
Assumption:
| Item | Estimate |
|---|---|
| Average conversation length | 3–5 AI replies |
| Average credit usage per query | 1.5 – 2.5 credits |
Monthly Cost Estimate
| Monthly Queries | Estimated Credits Used | Cost in USD | Cost in BDT | Cost in INR |
|---|---|---|---|---|
| 100 queries | 150 – 250 credits | $1.50 – $2.50 | ৳180 – ৳300 | ₹150 – ₹250 |
| 500 queries | 750 – 1,250 credits | $7.50 – $12.50 | ৳900 – ৳1,500 | ₹750 – ₹1,250 |
| 1,000 queries | 1,500 – 2,500 credits | $15 – $25 | ৳1,800 – ৳3,000 | ₹1,500 – ₹2,500 |
Example Use Cases
The AI Agent may answer questions such as:
- Is this product available?
- What is the price?
- Do you offer cash on delivery?
- How long does delivery take?
- What is your return policy?
- Can you recommend a product for me?
Example 2: SaaS Customer Support
A SaaS company uses Wize AI Agent to answer product usage, subscription, login, billing, and troubleshooting questions.
Assumption:
| Item | Estimate |
|---|---|
| Average conversation length | 5–7 AI replies |
| Average credit usage per query | 2.5 – 3.5 credits |
Monthly Cost Estimate
| Monthly Queries | Estimated Credits Used | Cost in USD | Cost in BDT | Cost in INR |
|---|---|---|---|---|
| 100 queries | 250 – 350 credits | $2.50 – $3.50 | ৳300 – ৳420 | ₹250 – ₹350 |
| 500 queries | 1,250 – 1,750 credits | $12.50 – $17.50 | ৳1,500 – ৳2,100 | ₹1,250 – ₹1,750 |
| 1,000 queries | 2,500 – 3,500 credits | $25 – $35 | ৳3,000 – ৳4,200 | ₹2,500 – ₹3,500 |
Example Use Cases
The AI Agent may help users with:
- Account login problems
- Subscription questions
- Feature explanation
- Billing-related questions
- Product setup guidance
- Troubleshooting steps
- Escalation to a live support agent when needed
How to Estimate Your Monthly AI Credit Usage
You can use this simple formula:
Monthly AI Credits = Number of monthly conversations × Average credits per conversation
Example:
If your business receives 500 AI-handled conversations per month, and each conversation uses around 2 credits, then:
500 × 2 = 1,000 AI credits per month
Then you can calculate the estimated cost:
| Market | Formula | Estimated Cost |
|---|---|---|
| Bangladesh | 1,000 credits × ৳1.20 | ৳1,200 |
| USA / USD Market | 1,000 credits × $0.01 | $10 |
| India | 1,000 credits × ₹1 | ₹1,000 |
Practical Monthly Planning Table
| Monthly Conversation Volume | Short Conversations | Normal Conversations | Complex Conversations |
|---|---|---|---|
| 100 conversations | 50 – 100 credits | 250 – 400 credits | 600 – 1,000 credits |
| 500 conversations | 250 – 500 credits | 1,250 – 2,000 credits | 3,000 – 5,000 credits |
| 1,000 conversations | 500 – 1,000 credits | 2,500 – 4,000 credits | 6,000 – 10,000 credits |
| 5,000 conversations | 2,500 – 5,000 credits | 12,500 – 20,000 credits | 30,000 – 50,000 credits |
Use this table as a planning guide. If your AI Agent mainly answers simple FAQs, your usage will stay closer to the lower range. If your AI Agent handles sales, support, troubleshooting, recommendations, or workflow actions, usage may move toward the higher range.
Best Practices to Manage AI Credit Usage
You can control and optimize AI credit usage by setting up your Wize AI Agent properly.
1. Train Only Useful Knowledge
Avoid training unnecessary pages, outdated content, duplicate blogs, or low-value pages. Start with the content customers actually ask about.
2. Keep Answers Clear and Direct
Long answers use more tokens. Configure your AI Agent to provide helpful but concise answers unless the customer asks for more detail.
3. Organize Your Knowledge Base
Use clean content, clear headings, updated FAQs, and properly written support articles. A well-structured knowledge base helps the AI Agent find the right answer faster.
4. Use Escalation Rules
Not every question should be handled by AI. For complex, sensitive, or high-value cases, you can set escalation rules so the AI Agent hands the conversation over to a live agent.
5. Review AI Performance Reports
Monitor AI engagement, resolution rate, unanswered questions, and usage trends. This helps you understand where credits are being used and how to improve your AI Agent setup.
6. Update Knowledge Regularly
If your pricing, policies, product details, or support process changes, update and retrain the knowledge base. Accurate content reduces repeated questions and unnecessary back-and-forth conversations.
Important Notes
- AI credit usage is estimated, not fixed.
- Actual usage may vary depending on customer message length, AI response length, knowledge base size, conversation history, and workflow complexity.
- Longer conversations naturally use more credits.
- Knowledge base training usually uses fewer credits compared to ongoing monthly conversations.
- Sales and troubleshooting conversations usually consume more credits than simple FAQ conversations.
- Clear AI Agent instructions can help reduce unnecessary long responses.
- Regular monitoring helps you control usage and improve performance.
Start free trial