Flexible endoscopy does not have a system that can automatically evacuate surgical smoke generated in the gastrointestinal lumen. We aimed to investigate the feasibility and potential usefulness of automatic smoke evacuation systems in flexible endoscopy. [Bench] After surgical smoke generated in the stomach was evacuated by the evacuator, the degree of residual smoke and gastric luminal collapse were evaluated to optimize the evacuator settings. [Animal] Insufflation, suction, and total operation time to complete the protocol of 10 cauterizations of the gastric mucosa were measured in three groups: "manual suction only," "manual suction with automatic evacuation (50% force)," and "manual suction with automatic evacuation (70% force)." The stability of endoscopic visualization and operability was evaluated by 10 endoscopists blinded to those suction settings, and the number of manual suctions, insufflations, and total operation time were measured. [Bench] The degree of residual smoke and gastric luminal collapse were inversely correlated. [Animal] When the automatic evacuator was partially used, there was no difference in the insufflation time, but the suction time (vs 50%; p = 0.011, vs. 70%; p = 0.011) and total operation time (vs. 50%; p = 0.012, vs. 70%; p = 0.036) were significantly reduced compared to manual operation only. Furthermore, manual suction with automatic evacuation (50% force) significantly improved the stability of endoscopic visualization and operability compared to manual operation only (p = 0.041, p = 0.0085). The automatic smoke evacuation in flexible gastrointestinal endoscopy was potentially feasible and useful by improving the device setting.