Skip to main content

Documentation Index

Fetch the complete documentation index at: https://quashbugs.com/docs/llms.txt

Use this file to discover all available pages before exploring further.

Every task you run in Quash consumes LLM API calls. The cost of a task depends on:
  • The complexity of the task (how much reasoning, validation and recovery is required)
  • The model used (for example, Claude or GPT)
  • The tools and reasoning capabilities enabled during execution
More complex tasks naturally require deeper reasoning, more decision-making and additional tooling, which in turn increases underlying API usage. Pricing scales with this complexity rather than with fixed or predefined steps. Quash does not charge based on:
  • Seats or users
  • Test cases or test suites
  • Projects or environments
  • Number of devices connected / parallel executions
Instead, pricing reflects the actual compute and API usage required to execute your task.

​
Free plan

When you sign up, you receive $10 in free usage credits. These credits let you:
  • Run real tasks
  • See how Quash breaks tasks down internally
  • Understand how pricing compounds with task complexity
There are no feature restrictions on the Free plan. Once your credits are exhausted, you can add funds and continue using Quash on a pay-as-you-go basis.

​
Pay-as-you-go pricing

Quash follows a task-based pay-as-you-go model. There are no subscriptions or fixed monthly fees. You are charged only when you run tasks. What you pay for Each task execution incurs:
  • LLM API usage cost (passed through from the model provider)
  • A ~20% platform markup to cover:
  • Agent orchestration
  • Execution reliability
  • Infrastructure and maintenance
  • Tooling, monitoring, and reporting
The platform markup is applied on top of the underlying API cost.

​
Why pricing varies between tasks

Tasks in Quash are not flat-priced. A task may:
  • Require multiple reasoning steps
  • Retry actions if the UI changes
  • Validate UI state or backend responses
  • Run longer depending on app behavior
As a result, more complex tasks consume more API calls, which increases cost. This is expected and intentional, so pricing scales with actual work done.

​
Example: how a task is priced

Consider the following task: “Open the app, log in with valid credentials, navigate to the profile screen and verify the username is displayed correctly.” To execute this, Quash may:
  • Interpret the task intent
  • Break it into multiple steps
  • Make several LLM API calls to plan and execute actions
  • Validate UI state and generate a report
Each of these steps consumes API calls. The total task cost is calculated by:
  1. Summing all LLM API usage during execution
  2. Applying the platform markup
  3. Deducting the final amount from your balance
A short, simple task may cost only a few cents. A longer or more complex task may cost more, depending on how many API calls are required.

​
Cloud device usage

Tasks can be executed on:
  • Local devices or emulators
  • Cloud-hosted devices
Free usage
  • Each account includes 60 minutes (1 hour) of cloud device usage at no cost, intended for evaluation and light testing.
Usage-based cloud devices
  • Cloud device usage beyond the free allowance is billed based on Cloud Device Minutes.
  • Usage is metered per minute, per device and clearly reflected in billing.
Dedicated cloud devices
  • You can request dedicated cloud devices assigned exclusively to your organization.
  • Dedicated devices are not usage-based and are billed separately.
  • Pricing is quoted based on:
    • Number of devices
    • Device models
    • Duration of reservation
  • Once provisioned, devices are reserved for your use for the agreed time period.
All cloud device costs are transparently tracked and documented.

​
Billing transparency

Quash provides detailed billing information for every task, including:
  • Task name
  • Execution status
  • Model used
  • Number of API calls
  • Total cost deducted
You can view this breakdown at any time in the Billing section under Profile.

​
Enterprise plans

Enterprise plans are designed for teams with advanced scale, security and infrastructure requirements. Enterprise plans include:
  • Unlimited usage across Agent Actions and test execution
  • Dedicated cloud devices assigned exclusively to your organization
  • Support for bringing your own LLM API keys (for example, via Azure OpenAI or Amazon Bedrock) to run Quash using your preferred models and accounts
Enterprise pricing is customized based on:
  • Device requirements (quantity and models)
  • Infrastructure and deployment preferences
  • Support, compliance and security needs
If you’re evaluating Quash for a larger organization, contact us to design a plan that fits your workflow.