Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Our enterprise solution with unlimited test execution, assisted setup, and dedicated support.
.claude/skills/testdriverai-testdriver-self-hosted/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-23 | ✗→✓ | ▲ Improved | 3% | 0% |
| case-01 | ✗→✓ | ▲ Improved | -18% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 16% | 0% |
| case-03 | ✗→✓ | ▲ Improved | 38% | 0% |
| case-07 | ✗→✓ | ▲ Improved | -12% | 0% |
<!-- Generated from self-hosted.mdx. DO NOT EDIT. -->
Self-hosted is our enterprise solution for teams that need unlimited test execution, infrastructure control, and dedicated support. Pricing is based on parallel test capacity with a flat license fee — no per-second billing.
<CardGroup cols={2}> <Card title="Unlimited Execution" icon="infinity"> Run as many tests as you want with no device-second metering. Predictable monthly costs. </Card> <Card title="Assisted Setup & Support" icon="headset"> Our team helps you deploy, configure, and optimize your infrastructure. Dedicated engineering support included. </Card> <Card title="Full Control" icon="gear"> Use your own AI keys, custom hardware, specific software, and network configurations. RDP into test machines for debugging. </Card> <Card title="Security & Compliance" icon="shield-check"> Keep data in your environment. Air-gapped deployment available for regulated industries. </Card> </CardGroup>
Choose the level of control you need:
| Component | Standard | Air-Gapped | |-----------|----------|------------| | Test Sandboxes | Your AWS | Your infrastructure (any cloud or on-prem) | | Dashboard | TestDriver hosted | Your infrastructure | | API | TestDriver hosted | Your infrastructure | | AI Processing | Your API keys | Your infrastructure | | Data Storage | Your AWS account | 100% your infrastructure | | Network | Internet access required | Fully air-gapped | | Cloud Providers | AWS | AWS, Azure, GCP, on-prem |
Run test sandboxes on your AWS infrastructure while using TestDriver's hosted dashboard and API:
<Card title="AWS Setup Guide" icon="aws" href="/v7/aws-setup" > Step-by-step instructions for deploying TestDriver on your AWS infrastructure using CloudFormation. </Card>
Deploy the entire TestDriver stack in your environment for complete isolation:
Build test environments with your applications, dependencies, and user data pre-installed. You get full access to:
<AccordionGroup> <Accordion title="What can you customize?">
</Accordion>
<Accordion title="How it works">
</Accordion> </AccordionGroup>
<Steps> <Step title="Discovery Call"> Discuss your requirements, security constraints, and integration needs with our team. </Step>
<Step title="Architecture Review"> Our engineers design a deployment architecture that meets your security and compliance requirements. </Step>
<Step title="Deployment"> We work with your team to deploy TestDriver, including assisted setup and configuration. </Step>
<Step title="Integration"> Connect TestDriver to your CI/CD pipelines, internal tools, and workflows. </Step>
<Step title="Training & Handoff"> Comprehensive training for your team on operating and maintaining the deployment. </Step> </Steps>
| Feature | Hosted | Self-Hosted | |---------|--------|-------------| | Setup Time | Minutes | Hours (assisted) | | Pricing Model | Device-seconds | Flat license fee | | Infrastructure | TestDriver | Your AWS or any cloud | | AI API Keys | TestDriver's | Your own | | Custom Software | Limited | Full control | | Hardware Selection | Standard | Your choice | | Debugging Access | Replays only | Full RDP access | | Support | Community/Standard | Dedicated engineering | | Air-Gapped Option | No | Yes |
<Card title="Schedule a Consultation" icon="calendar" href="https://testdriver.ai/demo" > Discuss your requirements with our team and get a custom proposal for your self-hosted deployment. </Card>
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-05 | pass→pass | 4,656 | 3,342 | -28% | 1 | 1 | 0% | 837 | 1,944 | +132% | 0 | 0 | — |
case-23 | fail→pass | 18,616 | 6,814 | -63% | 1 | 1 | 0% | 2,503 | 2,573 | +3% | 0 | 0 | — |
case-01 | fail→pass | 16,366 | 8,749 | -47% | 1 | 1 | 0% | 2,587 | 2,127 | -18% | 0 | 0 | — |
case-02 | fail→pass | 12,687 | 5,391 | -58% | 1 | 1 | 0% | 2,023 | 2,351 | +16% | 0 | 0 | — |
case-03 | fail→pass | 30,550 | 10,544 | -65% | 1 | 1 | 0% | 2,309 | 3,177 | +38% | 0 | 0 | — |
case-04 | pass→pass | 3,596 | 5,386 | +50% | 1 | 1 | 0% | 534 | 2,146 | +302% | 0 | 0 | — |
case-06 | pass→pass | 2,378 | 2,986 | +26% | 1 | 1 | 0% | 408 | 1,791 | +339% | 0 | 0 | — |
case-07 | fail→pass | 14,725 | 4,325 | -71% | 1 | 1 | 0% | 2,437 | 2,133 | -12% | 0 | 0 | — |
case-08 | pass→pass | 5,743 | 2,029 | -65% | 1 | 1 | 0% | 1,016 | 1,728 | +70% | 0 | 0 | — |
case-09 | pass→pass | 9,850 | 3,469 | -65% | 1 | 1 | 0% | 1,596 | 1,939 | +21% | 0 | 0 | — |
case-10 | pass→pass | 7,676 | 1,575 | -79% | 1 | 1 | 0% | 1,281 | 1,637 | +28% | 0 | 0 | — |
case-11 | fail→pass | 8,958 | 2,222 | -75% | 1 | 1 | 0% | 1,227 | 1,736 | +41% | 0 | 0 | — |
case-12 | pass→pass | 10,358 | 3,875 | -63% | 1 | 1 | 0% | 1,430 | 2,011 | +41% | 0 | 0 | — |
case-13 | fail→pass | 6,762 | 2,179 | -68% | 1 | 1 | 0% | 1,036 | 1,622 | +57% | 0 | 0 | — |
case-14 | fail→fail | 4,985 | 1,614 | -68% | 1 | 1 | 0% | 767 | 1,656 | +116% | 0 | 0 | — |
case-15 | fail→pass | 11,484 | 2,108 | -82% | 1 | 1 | 0% | 1,825 | 1,704 | -7% | 0 | 0 | — |
case-16 | fail→pass | 10,403 | 3,209 | -69% | 1 | 1 | 0% | 1,412 | 1,777 | +26% | 0 | 0 | — |
case-17 | pass→pass | 10,229 | 2,392 | -77% | 1 | 1 | 0% | 1,584 | 1,797 | +13% | 0 | 0 | — |
case-18 | pass→pass | 12,968 | 3,867 | -70% | 1 | 1 | 0% | 1,713 | 2,101 | +23% | 0 | 0 | — |
case-19 | pass→pass | 8,561 | 3,486 | -59% | 1 | 1 | 0% | 1,311 | 1,900 | +45% | 0 | 0 | — |
case-20 | fail→pass | 4,871 | 1,346 | -72% | 1 | 1 | 0% | 731 | 1,592 | +118% | 0 | 0 | — |
case-21 | pass→pass | 13,596 | 4,586 | -66% | 1 | 1 | 0% | 2,013 | 1,929 | -4% | 0 | 0 | — |
case-22 | pass→pass | 14,011 | 2,372 | -83% | 1 | 1 | 0% | 2,101 | 1,779 | -15% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 23 cases were attempted. The headline lift of +43 percentage points is the difference between those two pass rates over the 23 comparable cases.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.