Version 2.20.0
2025-12-01Model upload, the image benchmark type and a two-language interface.
New
- Models page: upload your own AI models and use them in benchmarks.
- Large model files are uploaded straight to storage.
- First version of LLM judge evaluation, question-answer benchmarks included.
- Invitation code management.
- Image benchmark type.
- A wizard to create multiple-choice benchmarks from an Excel file.
- Documentation in Turkish and English.
Improvements
- The interface is in Turkish and English; the calendar follows the selected language.