More than one in three couples grapple with relationship issues. Rather than seeking traditional couples therapy, couples are increasingly opting for online information and advice. Recent advancements in AI have led to chatbots, such as ChatGPT, capable of providing therapeutic responses. Early research evaluating a chatbots’ responses to relationship-related questions found they are perceived as more empathic than relationship therapists. Additionally, using simulated relationship therapy sessions, relationship therapists rated chatbots high on attributes such as empathy. However, such sessions did not use real participants, so we do not yet know how users perceive ChatGPT within a relationship therapy setting, nor have the technical outcomes been systemically evaluated. In the present study, 20 participants engaged in a single session of relationship therapy with ChatGPT and were interviewed about their experiences. We evaluated the performance of ChatGPT comprising of technical outcomes such as error rate and linguistic accuracy and therapeutic quality such as empathy and therapeutic questioning. The interviews were analysed using reflexive thematic analysis which generated four themes: light at the end of the tunnel; clearing the fog; clinical skills; and therapeutic setting. The analyses of technical and feasibility outcomes, as coded by researchers and perceived by users, show ChatGPT provides realistic single-session therapy with it consistently rated highly on attributes such as therapeutic skills, human-likeness, exploration, and useability, and providing clarity and next steps for users’ relationship problem. This study highlights the potential capabilities of ChatGPT in providing relationship advice and support.