Does a place-based openness reform change how firms in host cities collaborate on innovation with firms in other cities? We study this question for the inter-city co-patenting of cities hosting China's Pilot Free Trade Zones (PFTZs), which combine investment liberalisation, trade facilitation and administrative reform in designated areas. Using a panel of 296 prefecture-level cities for 2009-2022, we verify the treatment against State Council documents, which corrects the coding of six cities and yields 52 treated cities in six cohorts. Heterogeneity-robust staggered difference-indifferences estimates (Callaway-Sant'Anna, not-yet-treated controls), conditional on baseline development and co-patenting, associate designation with about 19% more utility-model co-patent counts (95% CI 2% to 39%) and 14% more co-patent counts of all types (CI-1% to 31%). These outcomes are counts rather than shares of total patenting. The utility-model result is the clearest across the four patent types but does not survive a Holm adjustment across those four overall estimates. The invention estimate is small and imprecise (6%, CI-8% to 22%); design shows no effect. The positive estimates disappear without baseline covariates and are not robust to extrapolating pre-period trends. Among cities with co-patents, the conditional response is concentrated in co-patents per partner rather than in the number of partners. IP-enforcement heterogeneity is not supported, and city-level data cannot identify whether PFTZs widened or narrowed regional co-patenting disparities. The paper contributes conditional evidence on openness and intercity collaboration, and shows how strongly conventional two-way fixed-effects estimates depend on the assumed counterfactual.