Charts from evaluating TypeSafe's Jev (a System One model) on Thai standardized exams, alongside 110 other language models. ภาษาไทยมีให้เลือกในแต่ละหน้า
Cost and speed vs accuracy 111 models on the same 375 questions, thinking models highlighted Confidence calibration Does Jev's stated confidence predict correctness? 567 questions, click any bin to read themOriginal data and results: ai-vs-thai-exams.pages.dev. Benchmark harness: dtinth/thaiexamjs. Exam questions from the ThaiExam dataset (SCB 10X) and the OpenThaiGPT evaluation dataset.