The increasing prevalence of multidrug-resistant tuberculosis (MDR-TB) highlights the urgent need for an efficient approach to identify Mycobacterium tuberculosis complex (MTBC) strains resistant to rifampicin (RIF) and isoniazid (INH). In response, we developed a CRISPR-Cas14a MTB RIF/INH platform that can detect the most common mutations associated with RIF and INH resistance. To evaluate the sensitivity and specificity of our CRISPR-Cas14a MTB RIF/INH platform, we carried out a comprehensive assessment using clinical isolates of M. tuberculosis and sputum samples from TB patients, making direct comparisons with phenotypic drug susceptibility testing (pDST). A total of 60 clinical isolates from TB patients were utilized, consisting of 18 RIF mono-resistant, 15 INH mono-resistant, 24 MDR isolates, and 3 fully susceptible isolates. Among the 42 RIF-resistant isolates, our platform accurately identified 39, achieving a sensitivity of 93.3% (95% CI, 80.0-98.5) and a specificity of 100% (95% CI, 81.6-100). Similarly, out of the 39 INH-resistant isolates, the platform successfully identified 38, demonstrating a sensitivity of 97.5% (95% CI, 86.5-99.9) and a specificity of 100% (95% CI, 83.8-100) when compared with pDST. Moreover, in the analysis of 55 sputum samples, our platform accurately identified RIF resistance in 10 out of 12 samples (85.7%) and INH resistance in all 11 samples (100%). Notably, excluding the nucleic acid extraction step, the entire testing procedure can be completed in approximately 1.5 h. These results suggest that the CRISPR-Cas14a MTB RIF/INH platform is a reliable and promising novel tool for detecting RIF and INH resistance in isolates or directly from sputum samples.