# Pricing & Usage

Understanding Chipp's usage-based pricing and how to optimize your AI costs


Understanding how AI pricing works helps you build better apps and control costs. This guide explains our pricing philosophy and gives you tools to estimate and optimize your usage.

## Why Usage-Based Pricing?

We get asked about pricing a lot, so let's be upfront about why we charge the way we do.

**We want to give every builder access to the best AI models available.** Not watered-down versions, not artificially limited options—the actual best models from OpenAI, Anthropic, and Google.

The challenge is that these models have real costs. Running GPT-5 costs more than running GPT-5 Nano. Claude Opus 4 costs more than Claude 3.5 Haiku. If we charged a flat rate, we'd have to either:

1. **Restrict access to premium models** — forcing everyone onto cheaper models
2. **Charge everyone the premium price** — making it unaffordable for many builders
3. **Lose money** — which means we can't stay in business and keep building tools for you

Usage-based pricing solves this. It means:

- **Budget-conscious builders** can use fast, cost-efficient models like Gemini 2.5 Flash or GPT-5 Nano and keep costs minimal
- **Quality-focused builders** can access the most powerful reasoning models like o3 or Claude Opus 4 when they need them
- **Smart builders** can mix and match—use affordable models for simple tasks, premium models for complex ones

This isn't about being greedy. It's about building a sustainable business that can keep providing best-in-class AI tools to every type of builder, at every budget level.

## How Pricing Works

### Your Plan Includes Usage

Every Chipp plan includes a monthly usage allowance:

| Plan | Monthly Price | Included Usage |
|------|---------------|----------------|
| Pro | $29/mo | $10 |
| Team | $99/mo | $30 |
| Business | $299/mo | $100 |

If you stay within your allowance, you pay nothing extra. If you need more, you can purchase additional credits in advance.

### What Counts as Usage?

Usage is measured in **tokens**—the units AI models use to process text. Roughly:
- 1 token ≈ 4 characters or ¾ of a word
- A typical message might use 500-1000 tokens total (input + output)

Different models have different costs per token. Premium models cost more, efficient models cost less.

Want to understand exactly how tokens work? Check out our [Understanding Tokens](/docs/guides/understanding-tokens) guide with an interactive tokenizer.

## Estimate Your Costs

Use this calculator to estimate your monthly AI costs based on expected usage:

<PricingCalculator />

## Model Pricing Comparison

Here's how different models compare on price. Choose based on your needs:

### Most Cost-Effective Models

These models offer excellent quality at the lowest prices:

<ModelComparison models={["gemini-2.5-flash-lite", "gemini-2.5-flash", "gpt-5-nano", "claude-3.5-haiku", "gpt-5-mini"]} />

### Premium Models

When you need the best quality, these models deliver:

<ModelComparison models={["gpt-5", "claude-sonnet-4", "claude-opus-4", "o3", "gemini-2.5-pro"]} />

## Strategies to Optimize Costs

### 1. Match Model to Task

Don't use a premium model for simple tasks. Consider:

| Task Type | Recommended Approach |
|-----------|---------------------|
| Quick Q&A, FAQs | Gemini 2.5 Flash or GPT-5 Nano |
| General conversation | GPT-5 Mini or Claude 3.5 Haiku |
| Content creation | GPT-5 or Claude Sonnet 4 |
| Complex analysis | o3 or Claude Opus 4 |
| Processing long documents | Gemini 2.5 Pro (best value for 1M context) |

### 2. Optimize Your Prompts

Shorter, clearer prompts use fewer tokens:

- **Be specific** about what you want
- **Remove unnecessary context** from system prompts
- **Use concise instructions** instead of verbose explanations

### 3. Consider Response Length

If your app doesn't need long responses, instruct the model to be concise. Output tokens typically cost more than input tokens.

### 4. Use the Right Context Window

Models with larger context windows (like Gemini 2.5 Pro with 1M tokens) are great for processing long documents. But if you're just doing quick Q&A, you don't need to pay for that capacity.

## Real-World Examples

### Example 1: Customer Support Bot

**Use case:** Answering common questions, routing complex issues

**Recommended:** Gemini 2.5 Flash ($0.25/M avg)

**Monthly cost for 10,000 messages:** ~$2.50

### Example 2: Content Writing Assistant

**Use case:** Helping users write blog posts, emails, marketing copy

**Recommended:** GPT-5 ($16/M avg)

**Monthly cost for 5,000 messages:** ~$40

### Example 3: Legal Document Analyzer

**Use case:** Reviewing contracts, extracting key terms, summarizing agreements

**Recommended:** Gemini 2.5 Pro ($4/M avg, 1M context)

**Monthly cost for 1,000 documents:** ~$20

### Example 4: Research Assistant

**Use case:** Deep analysis, multi-step reasoning, complex problem-solving

**Recommended:** o3 ($33/M avg)

**Monthly cost for 2,000 queries:** ~$33

## Monitoring Your Usage

You can track your usage in real-time:

1. Go to **Settings** → **Billing** in your Chipp dashboard
2. View your current credit balance and usage history
3. Purchase additional credits when needed

We'll automatically notify you when your credits are running low, so you have time to top up before your app pauses.

## FAQ

### What happens if I run out of credits?

Your app's AI functionality will pause until you purchase more credits. We'll send you notifications as your balance gets low so you're never caught off guard. You can top up anytime from your billing settings.

### How do I add more credits?

Go to **Settings** → **Billing** and purchase credits in advance. Credits are prepaid—there are no surprise bills or overages. You control exactly how much you spend.

### Do unused credits roll over?

No, included usage doesn't roll over month to month. But this keeps pricing simple and predictable.

### Why do output tokens cost more than input tokens?

This reflects how AI providers price their APIs. Generating new text (output) requires more computation than processing existing text (input).

### Can I change models mid-conversation?

Your app uses one model at a time, but you can change it anytime in the Build settings. The change applies to new conversations.

## What's Coming

Right now, usage billing covers chat interactions and tool calls. We're expanding this to unlock more powerful features for all builders:

### Voice Interactions

Real-time voice conversations with AI are expensive to run. By adding voice to usage billing, we can make this feature available to everyone—not just select customers. You'll pay for what you use, making voice accessible regardless of your plan.

### Knowledge Source Embeddings

Currently, bulk uploading knowledge sources is limited. Once we add embedding costs to usage billing, we'll be able to optimize for large-scale uploads. This means you'll be able to add hundreds or thousands of documents to power your app's knowledge base.

### Image & Video Generation

AI-generated images and videos are resource-intensive (video generation costs $3-6 per video). Usage billing lets us offer these creative tools to everyone. Generate marketing visuals, product demos, or custom content—paying only for what you create.

### Billing & Invoices

Your **Settings** → **Billing** page will show detailed usage invoices, so you can see exactly what you're paying for and track costs over time.

We're building toward a future where every AI capability is available to every builder, with transparent, pay-for-what-you-use pricing.

## Need Help Choosing?

If you're not sure which model fits your use case, check out our [Choosing a Model](/docs/guides/choosing-a-model) guide for detailed recommendations based on different scenarios.