A compound having the formula: L- (LU-D) p or a pharmaceutically acceptable salt or solvate thereof; wherein L is the ligand unit, LU is a linker unit and D is a drug unit, L is a peptide, a polypeptide or a protein that specifically binds to a population of target cells; LU has the formula -Aa-Ww-, Ww is a sequence of w selected independently of dirradical amino acids, which comprises a natural amino acid linked to an unnatural amino acid, wherein the W proximal to the drug unit (W1) is linked by a peptide bond to the drug unit, with the proviso that W1 cannot form a secondary amide with the amino acid of the C-terminus of D and that the peptide bond can be cleaved by an intracellular protease, w is an integer between 2 and 12, A is a widening unit, it is already 1 or 2; p is an integer from 1 to 20; and D has the formula: ** Formula ** in which the wavy line indicates the peptide bond with LU; R1 and R2 are independently selected from the group consisting of -H and C1-C8 alkyl, with the proviso that both R1 and R2 are not -H; R3 is selected from the group consisting of -H, -C1-C8 alkyl, C3-C8 carbocycle, -aryl, -C1-C8 alkyl aryl, -X1- (C3-C8 carbocycle), C3-C8 heterocycle and -X1- (C3-C8 heterocycle); R4 is selected from the group consisting of -H, -C1-C8 alkyl, -C3-C8 carbocycle, -aryl, -X1-aryl, -C1-C8 alkyl- (C3-C8 carbocycle), -C3-C8 heterocycle and -X130 - (C3-C8 heterocycle); R5 is selected from the group consisting of -H and methyl; or R4 and R5 together form a carbocyclic ring and have the formula - (CRaRb) n- in which Ra and Rb are independently selected from the group consisting of -H and -C1-C8 alkyl and n is selected from the group consisting of 2, 3, 4, 5 and 6; R6 is selected from the group consisting of -H and C1-C8 alkyl; R7 is selected from the group consisting of -H, -C1-C8 alkyl, C3-C8 carbocycle, -aryl, -X1-aryl, -X1- (C3-C8 carbocycle), -C3-C8 heterocycle and -X1- (C3-C8 heterocycle); each R8 is independently selected from the group consisting