> ## Documentation Index
> Fetch the complete documentation index at: https://browseruse-0aece648-mintlify-cli-docs-1773354647.mintlify.site/llms.txt
> Use this file to discover all available pages before exploring further.
# Models and Pricing
> Cloud API pricing, models, and cost breakdown.
## Pricing Overview
| Product | Pay As You Go | Business/Scaleup | Notes |
| -------------------- | ---------------------- | ---------------------- | --------------------------------------- |
| **AI Agent Tasks** | \$0.01 init + per-step | 25% off per-step | See LLM pricing below |
| **Browser Sessions** | \$0.06/hour | \$0.03/hour (50% off) | Charged upfront, unused time refunded |
| **Skill Creation** | \$2.00/skill | \$1.00/skill (50% off) | Refunded if cancelled, refinements free |
| **Skill Execution** | \$0.02/call | \$0.01/call (50% off) | Per API call |
| **Proxy Data** | \$10/GB | \$4-5/GB (50-60% off) | Metered by data transfer |
***
## AI Agent Task Pricing
Task pricing consists of two components:
1. **Task Initialization Cost**: \$0.01 per started task
2. **Task Step Cost**: Additional cost based on the specific model used for each step
### LLM Model Step Pricing
| Model | API String | Cost per Step |
| ------------------------ | ---------------------------- | ------------- |
| Browser Use 2.0 | `browser-use-2.0` | \$0.006 |
| Browser Use LLM | `browser-use-llm` | \$0.002 |
| O3 | `o3` | \$0.03 |
| Gemini 3 Pro Preview | `gemini-3-pro-preview` | \$0.03 |
| Gemini 3 Flash Preview | `gemini-3-flash-preview` | \$0.01 |
| Gemini Flash Latest | `gemini-flash-latest` | \$0.0075 |
| Gemini Flash Lite Latest | `gemini-flash-lite-latest` | \$0.005 |
| Claude Sonnet 4.5 | `claude-sonnet-4-5-20250929` | \$0.05 |
| Claude Sonnet 4.6 | `claude-sonnet-4-6` | \$0.05 |
**Browser Use 2.0** is the default — custom-trained for browser automation. 15x cheaper and 6x faster than frontier models (\~3s/step vs \~8s/step).
Use it by default (no configuration needed), or pass `llm` explicitly:
```python Python theme={null}
result = await client.run("...", llm="browser-use-2.0")
```
```typescript TypeScript theme={null}
const result = await client.run("...", { llm: "browser-use-2.0" });
```
### Example Cost Calculation
Using Browser Use 2.0 for a 10 step task:
* Task initialization: \$0.01
* 10 steps x \$0.006 per step = \$0.06
* **Total cost: \$0.07**
Using Browser Use LLM for a 10 step task:
* Task initialization: \$0.01
* 10 steps x \$0.002 per step = \$0.02
* **Total cost: \$0.03** (\~33 tasks per dollar)
***
## V3 API Pricing (Token-Based)
V3 uses token-based billing instead of flat per-step costs.
| Model | Input per 1M | Output per 1M |
| -------------------------- | ------------ | ------------- |
| BU Mini (Gemini 3 Flash) | \~\$0.72 | \~\$4.20 |
| BU Max (Claude Sonnet 4.6) | \~\$3.60 | \~\$18.00 |
Includes a 1.2x markup over raw provider costs. Task initialization (\$0.01) and browser sessions are billed separately.
***
## Browser Session Pricing
Browser sessions (direct CDP access) are billed separately from AI agent tasks. **Business and Scaleup subscribers get 50% off browser session costs.**
| Plan | Hourly Rate | Per Minute |
| ---------------- | ----------- | ------------ |
| Pay As You Go | \$0.06/hour | \$0.001/min |
| Business/Scaleup | \$0.03/hour | \$0.0005/min |
| Item | Details |
| -------------- | -------------------------------------- |
| Minimum charge | 1 minute |
| Billing | Charged upfront |
| Refunds | Proportional refund when stopped early |
### How Browser Session Billing Works
1. **Upfront charge**: Full amount charged when session starts based on timeout duration
2. **Proportional refund**: Unused time refunded when you stop the session
3. **Minimum billing**: At least 1 minute is charged
**Example (Pay As You Go)**: You start a 60-minute session (\$0.06 charged). You stop after 30 minutes. You receive a \$0.03 refund.
**Example (Business)**: Same scenario costs only \$0.03 upfront, with \$0.015 refund for unused time.
### Timeout Limits
All plans support up to **4 hours (240 minutes)** per browser session.
***
## Skill Pricing
| Item | Pay As You Go | Business/Scaleup |
| -------------------------- | ------------------- | ----------------------------- |
| Skill creation | \$2.00 per skill | \$1.00 per skill (50% off) |
| Skill execution | \$0.02 per API call | \$0.01 per API call (50% off) |
| Skill refinement | Free | Free |
| Minimum balance to create | \$2.00 | \$2.00 |
| Minimum balance to execute | \$0.10 | \$0.10 |
### How Skill Creation Billing Works
1. **Upfront charge**: Charged when you start creating a skill
2. **Refund on cancellation**: Full refund if you cancel before generation completes
3. **Refinements are free**: Iterating on an existing skill costs nothing
***
## Proxy Pricing
Proxy usage is charged per GB of data transferred. **Business and Scaleup subscribers get 50% off proxy costs.**
| Plan | Proxy Cost |
| ------------- | ---------------- |
| Pay As You Go | \$10/GB |
| Business | \$5/GB (50% off) |
| Scaleup | \$4/GB (60% off) |
Proxy is automatically enabled when you set `proxyCountryCode` on a session. Data usage is metered and charged based on actual transfer.
***
**Scaleup & Enterprise** — [contact us](https://browser-use.com/pricing) to get started. We sign ZDR (zero data retention), no-training, HIPAA, BAA, and custom compliance contracts. Includes dedicated browser pool, on-prem deployment options, and custom SLAs.
[See plans & pricing →](https://browser-use.com/pricing)