Scientific publications and supplementary data are a critical source of variant evidence, but they remain largely disconnected from GA4GH standards and ecosystem. This session will explore the integration of literature variant extraction pipelines with GA4GH standards to improve the discoverability, representation, provenance and supporting evidence of variant knowledge. We will discuss challenges in variant extraction with AI tools, normalisation into the Variation Representation Specification (VRS) and Cat-VRS, and the representation of associated data (e.g. diseases, phenotypes) and metadata (e.g. publication IDs) with the Variant Annotation Specification (VA-spec). We will also consider the deployment of a Beacon instance to expose the normalised literature variant corpus. The aim is to identify challenges and solutions, improve alignment with Cat-VRS and VA-Spec, explore the development of a global Variant-Beacon network and find collaboration opportunities.