Growing demand for pavement maintenance promotes reclaimed asphalt pavement (RAP) recycling, yet multi-scale comparative research on rejuvenators remains limited. This study compared a commercial bio-based Rejuvenator A and a petroleum-based Rejuvenator B to reveal how chemical compositions control asphalt regeneration. Following JTG 3410-2025 standards, asphalt binders underwent extended RTFOT aging (75, 112.5, and 150 min) and were externally blended with rejuvenators at 5%, 7.5%, and 10% by binder mass. A multi-scale framework combining macro tests, interfacial characterization, SARA fraction analysis, ATR-FTIR, and molecular dynamics (MD) simulations was established. Results indicated that the two rejuvenators acted via different mechanisms. Rejuvenator B contained 31.64% saturates and restored aged asphalt mainly through physical dilution and softening. This rejuvenator formed 0–80 hydrogen bonds dominated by weak C–H···O interactions, and the corresponding rejuvenated asphalt yielded a CII of 0.293. It was effective for 75 min-aged binders yet produced more oxidation products after re-aging. Conversely, Rejuvenator A, characterized by 68.10% aromatics and 20.17% resins, exhibited a 3–5° lower equilibrium contact angle, and a roughly 13% shorter relative penetration time than Rejuvenator B. Its network of 150–250 hydrogen bonds effectively disrupted asphaltene aggregates, reducing the CII to 0.180. Consequently, Rejuvenator A successfully restored 112.5 min- and 150 min-aged asphalt with significantly lower secondary oxidation. These performance differences were governed by aromatic and polar fractions rather than feedstock origins. While quantitative findings are sample-specific, the established framework provides theoretical references for engineering rejuvenator selection and high-performance regenerant design.