RNA length and receptor usage define innate immune recognition across species.
Recognition of viral double-stranded RNA (dsRNA) is a central trigger of innate antiviral immunity, yet the role of RNA length in shaping immune sensing across species remains poorly defined. Here, we identify RNA length as a key determinant of innate immune activation and demonstrate that this parameter is differentially tuned across vertebrate hosts. Using defined dsRNA molecules, we compared immune responses in human, duck, and chicken cells. Short 5-triphosphorylated dsRNAs, including influenza-derived mini viral RNAs, robustly activated antiviral responses in human and duck cells but failed to do so in chicken cells. In contrast, longer dsRNAs induced immune activation across all species, revealing a species-specific threshold for dsRNA recognition. Mechanistically, this threshold is defined by differential engagement of RNA sensors. In human cells, sensing of short and intermediate dsRNA is strictly dependent on RIG-I, whereas in chicken cells, which lack RIG-I, MDA5 mediates dsRNA recognition and imposes a higher length requirement for activation. Consistent with this, duck cells, which retain RIG-I, respond to short dsRNA similarly to human cells. These differences in dsRNA sensing are accompanied by differential activation of downstream antiviral pathways, including PKR-dependent translational control and OAS/RNase L activation. Together, our findings identify RNA length as a fundamental parameter governing innate immune recognition and reveal that species-specific usage of RNA sensors constrains antiviral sensing, with implications for host susceptibility and virus-host interactions. Significance StatementRecognition of viral RNA is central to antiviral immunity, yet the rules governing which viral RNAs are sensed across species remain poorly understood. Here, we show that RNA length defines a species-specific threshold for innate immune recognition. Human and duck cells efficiently detect short viral double-stranded RNAs generated during influenza infection, whereas chicken cells respond only to substantially longer RNA duplexes. Mechanistically, this difference reflects the usage of distinct antiviral RNA sensors: RIG-I enables sensing of short viral RNAs, while MDA5 imposes a higher length requirement for activation. Our findings reveal that host species differ in the spectrum of viral RNA ligands accessible to immune detection, providing a framework for understanding species-specific antiviral responses and host susceptibility.