The objective of this experiment was to determine the effect of increasing doses of a phytogenic product based on condensed tannins and spices (CTS) on production performance of lactating dairy cows fed a low protein diet. Eight rumen-cannulated Holstein Friesian dairy cows (140 ± 86 DIM; 39.0 ± 5 kg/d milk yield; mean ± SD), were used in a replicated 4 × 4 Latin Square design experiment with 4-wk periods. Treatments were: 0, 10, 20 and 30 g/d CTS (CTR, 10CTS, 20CTS, and 30CTS, respectively). The grass silage and corn silage-based diet was 55.2% forage, 38.7% NDF, 21.0% total starch, and 14.6% CP. Orthogonal contrasts were used to evaluate the linear and quadratic effect of increasing doses of CTS. Results follow the order: CTR, 10CTS, 20CTS, and 30CTS. Increasing doses of CTS quadratically increased DMI (25.4, 25.9, 26.1, and 25.1 kg/d) and milk yield (37.1, 38.5, 37.7, and 36.3 kg/d), tended to increase fat-and-protein-corrected milk (36.9, 37.6, 37.4, and 36.1 kg/d), and did not affect feed or N efficiency (1.45 ± 0.2 and 32.0 ± 2.3%, respectively). Treatments did not affect milk fat yield (1.48 ± 0.2 kg/d) but increasing doses of CTS increased milk protein yield quadratically (1.22, 1.27, 1.26, and 1.20 kg/d). Intermediate doses of CTS tended to increase de novo fatty acid yield (352, 369, 373, and 356 g/d) and decrease trans-10 C18:1 (4.31, 4.05, 4.05, and 4.24 g/d) compared with CTR and 30CTS. Treatments did not affect milk urea concentration (17.8 ± 1.7 mg/dL) or milk crude protein (3.39 ± 0.2%) or fat (4.06 ± 0.2%) content. Rumen pH and time below rumen pH of 5.8 were not affected by level of CTS supplementation. A treatment by time interaction for rumen ammonia concentration indicated that 20CTS and 30CTS increased ammonia concentration 3 h post-feeding compared with CTR and 10CTS (7.72, 7.94, 13.7, and 14.1 mg/dL). The 10CTS treatment decreased rumen propionate concentration only at 3 h post-feeding compared with the other treatments. Apparent DM and NDF total-tract digestibility were not affected by treatments. Shotgun metagenomics were used to evaluate the impact of CTS supplementation on the solid- and liquid-associated rumen microbiome. Treatment effects were only observed in the solid-associated microbiome. Supplementation of CTS linearly decreased α diversity at both the taxa and functional levels, indicating promotion of a leaner microbial community with higher doses of CTS. Differential abundance analysis identified 26 species with large fold changes, including some species with a high presence of cellulases and significant correlations with phenotypic parameters such as DMI, N efficiency, and milk production. In conclusion, a mixture of CTS affected microbiome and rumen metabolism, increasing fat-and-protein-corrected milk yield when fed at 10 and 20 g/d only. This experiment demonstrates the importance of in vivo dose response experiments with phytogenic products to determine optimum dosage for improved rumen metabolism and performance.