Short circulation times and off-target toxicity of toxin-based therapeutics typically limit their efficacy, highlighting the need for advanced delivery systems. Here, we employed protein cage nanoparticles (AaLS) as toxin delivery nanoplatforms to enhance blood circulation time and created a cancer-targeting toxin module (DTA-HER2Afb) by genetically fusing the A fragment and T-domain of diphtheria toxin (DTA) with the HER2-targeting affibody (HER2Afb). Multiple DTA-HER2Afbs were displayed on the AaLS surface using the SC/ST protein ligation system to form AaLS/DTA-HER2Afb, preserving stable, uniform nanoscale architectures. AaLS/DTA-HER2Afb selectively bound to HER2-overexpressiong SKOV3 cancer cells with minimal nonspecific binding. Both monomeric DTA-HER2Afb and AaLS/DTA-HER2Afb selectively delivered DTA to SKOV3 cancer cells with similar efficiency in vitro, effectively reducing cell viability irrespective of their configuration. However, AaLS/DTA-HER2Afb exhibited superior tumor accumulation at the tumor sites and more effectively suppressed tumor growth compared to the monomeric DTA-HER2Afb in SKOV3 tumor-bearing mice. The improved in vivo efficacy may be due to the nanoscale properties of AaLS/DTA-HER2Afb, which prolonged blood circulation time and enhanced tumor-specific accumulation via the EPR effect. Our study presents a promising approach for developing effective and safe toxin delivery nanoplatforms by combining a target-specific toxin delivery module with protein cage nanoparticles. Statement of significance: Short circulation and off-target toxicity often limit the therapeutic use of toxins, requiring improved delivery systems. Here, we designed a nanoplatform using self-assembling protein cage nanoparticles (AaLS) for targeted toxin delivery. A cancer-specific toxin module (DTA-HER2Afb) was engineered by fusing diphtheria toxin domains with a HER2-targeting affibody. Displayed on AaLS, the resulting AaLS/DTA-HER2Afb nanoparticles selectively targeted HER2-positive SKOV3 cells and delivered the toxin effectively. While in vitro efficacy was similar to monomers, the nanoparticles showed enhanced tumor suppression in vivo, benefiting from prolonged circulation and tumor accumulation via the EPR effect. The current study highlights the potential of combining target-specific toxin delivery modules with protein cage nanoparticles to offer a promising, safe in vivo platform for targeted cancer therapy.