A stimulated Raman scattering (SRS) technique-guided method for visualizing and evaluating in vitro cellular inflammation models is proposed, which provides new ideas for drug evaluation. A single-cell feature extraction process is established to obtain a total of 10 features including the morphology and composition of cells, and cells are segmented using spectral phasor analysis extraction to obtain the spectral information of different cellular regions. To validate the feasibility of the described method, an inflammatory cell model is constructed. Different mass concentrations of lipopolysaccharide (LPS) are used to induce phenotypic changes in RAW264.7 cells, accompanied by varying degrees of inflammatory responses. Results show that the cell area of the LPS-treated group is significantly larger (P<0.01) compared with the control group, and the ratio of intensity between 2855 cm^-1 and 2940 cm^-1 wave number in LD spectra is significantly increased (P<0.0001) and correlated with the mass concentration of LPS. The proposed method can realize the accurate extraction of single-cell feature information, which can be successfully applied to the differentiation of LPS-induced models with different degrees of inflammation, and verifies the applicability of SRS technology in the visual assessment of in vitro cellular models.