What's new in Explyt Test 3.1
Explyt Test 3.1 takes a significant step forward in delivering an even better test generation experience introducing revised prompts algorithms, incremental fixes that do not rewrite the whole test file, and GPT-4.1 support.
A range of enhancements driven by community feedback, improve code quality and make test writing less time-consuming and more comfortable for developers.
Reproduce, inspect and fix inside JetBrains IDEs, with the debugger, run configurations and IDE facts as evidence.
Compatibility updates:
- supported custom OpenAI-compatible large language model providers
- OpenAI GPT-4.1, o4-mini, o3
- Gemini 2.5
- Claude 4
- Mistral Devstral
Key updates:
- significantly improved and accelerated mechanism for applying LLM fixes
- saving user prompts for the project
Overall improvements:
- revised prompts for mocking help achieve more precise mocks
- revised prompts to follow reference test class and additional prompts better
- detect and prune duplicate generated tests when increasing test coverage
Minor updates and bug fixes:
- improved context collection algorithm
- fixed Undo/Redo integration issues in IDE
- fixed Windows-specific instrumentation (test runner) bugs
- supported setting package name when generating tests
- supported Gradle test root creation (other build systems have been supported earlier)
- enhanced Add Attachments mechanism
- further improved Coverage Viewer accuracy
Explyt Test supports test frameworks: JUnit 5, JUnit 4, TestNG, kotlin-test, mocking frameworks: Mockito, Mockito-Kotlin, MockK, SpringMockK, and build systems: Maven, Gradle, IntelliJ Build System.
You can access language models either through Explyt servers or by using your own API-keys.
Explyt Enterprise with Virtual Private Cloud (VPC) or self-hosting solution guarantees corporate data security. Your company's code is not sent to the internet and is not used for model training.
Reproduce, inspect and fix inside JetBrains IDEs, with the debugger, run configurations and IDE facts as evidence.


