AI & Advanced
Leverage AI to generate test cases, get insights, connect AI agents, and power advanced reports.
Overview
TestKase's AI features use large language models to accelerate your testing workflow. Test Case Generation and the AI Agent make real LLM calls to generate test cases from simple descriptions and answer questions about your project. The Insights reports, by contrast, use advanced analytics and heuristic scoring computed directly from your data. Together these capabilities help your team work smarter and ship higher-quality software.
AI in TestKase is designed to augment, not replace, your testing expertise. Every AI-generated output should be reviewed and validated by your team before being used in production testing.
Test case generation
Draft test cases from a pasted requirement or a linked one
AI Agent
Built-in conversational AI for test management
11 insight reports
Advanced analytics for flaky tests, risk, quality & more — no credits used
AI credits
One 100-credit grant on Free; 100 per seat every month on a paid plan. Generation & chat are token-based
Org-level control
On by default; the owner can disable AI for the whole organization
Private & secure
Data encrypted in transit, not used for AI training
AI Credits
AI features that call a language model (Test Case Generation and the AI Agent) are powered by a credit system. Credits are consumed based on the number of tokens each request uses. Credits belong to the organization and are shared across all its users.
How Credits Work
- Shared balance — Credits belong to the organization, not individual users. All members share the same pool.
- One-time grant of 100 credits on Free — A free organization receives 100 credits when it is created, for the life of the account. They are never topped up; once spent, AI features stop until the organization upgrades.
- Monthly grant of 100 credits per seat on paid plans — The rate is the same on Premium and Enterprise, so the balance is simply 100 × your seat count, refreshed on the 1st of every month.
- Not carried over — On a paid plan unused credits do not roll over; the balance is overwritten, not accumulated. The balance is not top-up purchasable on any plan.
- Token-based consumption — Generation and AI Agent messages are charged by the actual number of tokens used, not a flat per-request fee.
- Visible balance — The current credit balance is shown on the Organization → AI Credits page and in the AI credit indicator in the navigation bar.
What Consumes Credits
| AI Feature | Credit Cost | Notes |
|---|---|---|
| Test Case Generation | Token-based | Scales with description length and the size of the generated output. One request returns at most 16,384 tokens, whether you asked for 1 test case or 10 |
| AI Agent message | Token-based | Scales with the size of your prompt, project context, and the model's response. A message may take up to 10 tool-calling rounds, each re-sending the conversation so far |
| Insights reports (Tier 3) | None | Computed from your data with SQL analytics — they do not consume credits |
| Score badges | None | Flakiness and Duplicate scores do not consume credits |
On a paid plan your monthly allocation is determined by your seat count alone — 100 credits per seat, at the same rate on Premium and Enterprise. Upgrading does not raise the rate; only adding seats raises the balance. Because the balance is refreshed on the 1st of each month, plan seats up front for the AI usage your team expects rather than relying on carry-over.
A free organization is outside this cycle entirely: it receives 100 credits once, for the life of the account, and nothing is added on the 1st.
AI Controls for Your Org
AI features are enabled by default for a new organization — there is no switch to turn on before your team can use them. To review or change that:
- Navigate to Organization → AI Credits (
/admin/ai-credits). - Toggle Enable AI Features.
Only the organization owner can change the toggle. Everyone in the org can open the page and read the balance.
What Happens When AI Is Enabled
- Create with AI appears in the New dropdown on the Test Cases page.
- The AI Agent sparkle button is active in the top navigation bar.
- A credit balance indicator appears in the navigation bar.
What Happens When AI Is Disabled
- All AI-related UI elements are hidden across all projects.
- No credits are consumed (features cannot be triggered).
- Previously generated AI content (test cases, chat history) is preserved.
- The Insights reports and the Flakiness/Duplicate score badges keep working — they use no AI.
You can enable and disable AI at any time without losing any data. Use this toggle to control costs during periods when AI is not needed.
AI Quality Level
Each AI feature (the AI Agent and Test Case Generation) has an independent quality level — Basic, Balanced or Premium — set on the same AI Credits page. Both start on Balanced, and the level changes how much a call costs by a wide margin:
| Quality level | AI Agent | Test Case Generation |
|---|---|---|
| Basic | ~34,000 tokens per credit | ~22,500 tokens per credit |
| Balanced (default) | ~8,500 tokens per credit | ~3,000 tokens per credit |
| Premium | ~2,850 tokens per credit | ~2,850 tokens per credit |
That is roughly a 12x spread for the AI Agent and 8x for generation. If your balance is disappearing faster than you expect, check the level before you add seats. See What an AI Action Costs and AI Privilege & Quality Settings.
Score Badges
Every test case detail view shows two score badges — Flakiness Score and Duplicate Score — directly in the tab bar. These give you at-a-glance quality signals without running a full report.
- Flakiness Score measures how often a test case changes result across executions. Pass, fail and blocked all count as results; resetting a case to not-executed does not. High scores indicate unreliable tests that need attention.
- Duplicate Score measures how closely a test case title matches other titles in your project. High scores suggest redundancy in your suite.
Hover over either badge to see a detailed breakdown. Badges are color-coded (green/blue = healthy, yellow = review, red = action needed) and do not consume AI credits.
See Test Cases → Score Badges for full details on score ranges and color thresholds.
Next Steps
- AI Test Case Generation -- Generate comprehensive test cases from natural language descriptions.
- AI Agent -- Built-in conversational AI sidebar for creating test cases, querying data, and managing tests through natural language.
- Insight Reports -- Advanced analytics including flaky tests, risk heatmaps, and predictive failure.
- MCP Server -- Connect AI agents like Claude, Copilot, and Cursor to TestKase.
- Credits & Billing -- Understand the monthly credit grant, monitor usage, and data privacy.
- Best Practices & FAQ -- Guidelines for effective AI usage and common questions.