Pulsed fiber lasers have attracted much attention owing to many potential applications in such areas as environmental sensing, biomedical diagnostics, and nonlinear frequency generation.1,2 Mainly two schemes are selected for generating pulsed laser emission—active mode-locking or Q-switching by employing costly and complex electrically driven modulators,3–5 and passive mode-locking or Q-switching by incorporating an artificial or real saturable absorber (SA) as an intensity-dependent optical switch. Compared with active scheme, passive scheme has the advantages of compactness, simplicity, and flexibility. In the past, the nonlinear polarization rotation (NPR), nonlinear optical loop mirror (NOLM),6 and semiconductor saturable absorber mirror (SESAM)7 were the dominant techniques with wide application in commercial laser system. Both NPR and NOLM are the fiber nonlinearity-based artificial SAs with higher damage threshold, but are also highly sensitive to environmental fluctuations. SESAM requires complicated fabrication and packaging process, and has limited bandwidth, few picoseconds response times. From the last decade, carbon nanotube (CNT)8–15 has emerged as the first case of nanomaterial SA for mode-locking, which has one-dimensional (1-D) nanostructure and owns advantages such as ultrafast photoresponse, easy fabrication, and low cost. Gold nanorod (GNR)16,17 is another familiar 1-D nanomaterial SA, which has two surface plasmon resonance absorption bands and has been utilized for Q-switching or mode-locking. These 1-D nanomaterial SAs have absorption peaks related with the tube diameter (for CNT) or the aspect ratio (for GNR), thus leading to relatively narrowband operation restricted by the optimal absorption band. For wideband operation, the 1-D nanomaterial SAs require the combination of their nanotubes with a broad diameters distribution or nanorods with different aspect ratio. Unlike CNT or GNR, graphene18,19 has a Dirac-like electronic band structure with unique zero bandgap, which endows it with remarkable optical properties (i.e., ultrafast photoresponse, ultrabroadband absorption) and electrical properties (i.e., remarkable electron mobility). Extensively, research20–47 has employed graphene as a candidate material for the development of functional photonic technologies, e.g., ultrafast mode-locker, broadband polarizer, and phase shifter. But graphene also holds two main disadvantages, the weak modulation depth (typically per layer20) and the difficulty of creating an optical bandgap. Therefore, substantial endeavors have been focused on developing new SA beyond graphene from other layered crystal,48 such as topological insulators (TIs),49–82 transition mental dichalcogenides (TMDs),83–114 including of molybdenum disulfide () or tungsten disulfide (), as well as their diselenide analogues (, , , and ) and black phosphorus.115–118 The series of TIs (i.e., , , , and so on) demonstrate the characteristics of a small bandgap in the bulk state and a gapless metallic state in the edge/surface. Both the bulk- and nanostructured-TIs have been applied in pulsed laser cavity as the SA device. The TMDs possesses a thickness-dependent electronic and optical property—their bulk states have indirect bandgap with weak light–matter interaction while monolayer or few-layer structures are direct bandgap semiconductor with enhanced light activity. For example, a bulk is a semiconductor with bandgap of 1.4 eV (), while its monolayer has a direct bandgap of 2.1 eV (). Its thickness-dependent bandgap and electronic band structure endow it with many potential optical applications in optical fields that require strong light–matter interaction. These newly emerged two-dimensional (2-D) materials possess remarkable abilities in the field of optoelectronics and nonlinear photonics, such as broadband SA, high third-order nonlinear susceptibility, and ultrafast carrier dynamics.