Cadillac XT5 PHEV to Debut at Chengdu Auto Show, Featuring Momenta R7 World Model

The Cadillac XT5 PHEV is expected to officially enter pre-sale at the Chengdu Auto Show, opening on August 21, 2026. As the first plug-in hybrid variant of the XT5 lineup, this mid-size SUV introduces the Momenta R7 world model intelligent driving system, along with LiDAR sensors and an Ultium-platform nickel-cobalt-manganese (NCM) battery — marking a significant step for Cadillac in integrating intelligence and electrification.

Exterior design retains the overall silhouette and signature language of the current gasoline-powered XT5: a shield-shaped front grille paired with a mesh insert, and split LED headlamps enhancing visual distinctiveness; LiDAR is integrated into the roof to provide hardware redundancy for advanced intelligent driving. Side profile features conventional door handles and multi-spoke wheels, while gloss-black window trim elevates refinement; the rear adopts T-shaped LED taillights complemented by a full-width chrome garnish, an upward-swept rear spoiler adding dynamism, and a concealed exhaust layout replacing the previous dual-outlet setup. Overall dimensions measure 4888/1957/1685 (1688) mm (L/W/H), with a wheelbase of 2863 mm.

Powertrain comprises a 1.5T engine and front/rear dual electric motors: the engine delivers peak output of 115 kW, while the front and rear motors produce up to 160 kW and 110 kW respectively. Paired with a 35.5 kWh Ultium 5C NCM battery, the vehicle achieves a CLTC all-electric range of 155 km and a WLTC combined fuel consumption of 6.05 L/100 km. The chassis features the all-terrain electric Hummingbird Chassis, engineered for superior adaptability across diverse road conditions and enhanced dynamic responsiveness.

On the intelligent driving front, the XT5 PHEV will be the first production model to feature the Momenta R7 world model. Unlike traditional perception-based ADAS systems, the R7 model's core strength lies in physical-world modeling — it understands object materials, motion causality, and interaction possibilities, enabling more physically plausible trajectory prediction and avoidance strategies in scenarios such as rush-hour commuting, narrow-road passing, trace-reversing, complex intersection navigation, and automated parking. This capability remains in functional preview stage; real-world performance awaits verification through on-road testing.