Verke Editorial · Last verified: 2026-04-19

Verke vs. Therabot: a consumer coaching product next to a Dartmouth research artifact

Same evidence trajectory, different stage of the journey.

Therabot is the AI mental-health product that put generative coaching on the clinical map — a Dartmouth-led randomized controlled trial published in NEJM AI in March 2025, covered widely by MIT Technology Review, showing meaningful symptom reductions across depression, generalized anxiety, and eating-disorder concerns over an eight-week treatment period. It is also not a product you can sign up for. Therabot exists as a research artifact at the Geisel School of Medicine; there is no consumer app, no pricing page, no announced public launch. Verke is an AI coaching app you can actually use today, built on the same conviction that purpose-built generative AI grounded in evidence-based methods can help — and the subject of its own published randomized controlled trial. In the Stockholm University trial, participants often described being able to reach their coach in the moment they needed it, getting concrete tools to try, and coming to understand their own patterns.

Verke is an AI coaching app with five specialist coaches trained on evidence-based therapeutic methods (CBT, PDT, ACT, EFT, CFT, NVC). Chat in text or switch to voice for a phone-call-style session; pick up again days later with memory that survives across weeks and months. Conversations are encrypted, you can start anonymously, and the product is available on iOS, Android, and Web in 55 languages.

The five coaches are differentiated on purpose. Anna works in a psychodynamic register — slower, pattern-first, interested in why the same kind of situation keeps returning. Judith is CBT for social anxiety — small experiments, predictions tested, gradual exposure. Marie supports couples through Emotionally Focused Therapy and Nonviolent Communication; Amanda does general CBT, with Acceptance and Commitment Therapy and Compassion-Focused Therapy for overwhelm and self-criticism; Mikkel is an executive coach for decision fatigue and leadership load. Users pick who fits the concern, rather than asking one bot to cover everything.

A Verke session is shaped around two things: being able to reach a coach in the moment something feels acute rather than waiting for an appointment, and having an overwhelming problem broken down into concrete next steps small enough to actually do. The coach mirrors what you bring without performance pressure, and over multiple weeks it begins to connect threads — the recurring trigger, the avoidance loop, the strategy that has quietly started to work — so the work accumulates instead of resetting.

Therabot is a research-only generative AI mental-health tool built at Dartmouth's Geisel School of Medicine. It became the first generative AI mental-health intervention to publish randomized controlled trial results, in NEJM AI in March 2025, with the study covered in MIT Technology Review (March 2025). The trial enrolled 210 adults with significant symptoms of major depression, generalized anxiety, or eating-disorder risk; participants used Therabot for eight weeks and showed roughly 51% reductions in depression symptoms, 31% in GAD symptoms, and 19% in eating-disorder and body-image concerns. Principal investigator Nick Jacobson personally oversaw all participant messages during the trial — a research design choice that does not scale to a consumer launch. Therabot has not been released as a public product, has no pricing, and there is no announced commercial availability.

You can't — not as a consumer. Therabot is a research artifact, and there is no published roadmap for a public launch. If the Therabot RCT's results resonated with you and you want to engage with AI-delivered coaching grounded in the same evidence-based traditions, Verke is the closest counterpart you can actually use. We're on the same evidence trajectory at a different stage of the journey: Therabot is still a research artifact; Verke has the live consumer product, a published academic RCT at Stockholm University, and a roster of specialist coaches you can start a conversation with right now.

Verke is built for people who want a coach matched to a specific concern rather than a generalist, and who want to start today. If you're working with anxiety in a structured CBT register — the closest concern in Verke's roster to Therabot's research conditions — Amanda is the right starting point. For social anxiety there's Judith; for old relational patterns, Anna; for couples work, Marie; for high-pressure leadership, Mikkel. Voice coaching in a phone-call format is core to the product, conversations are encrypted in transit and at rest, and two of Verke's coaches were evaluated in a published Stockholm University trial supervised by Professor Per Carlbring. Read the method explainer for Cognitive Behavioral Therapy — the modality with the largest overlap with Therabot's training base.

Two practical differences are worth calling out. First, memory: Verke uses a three-tier memory system so the coach remembers what you've been working on weeks later without you re-explaining yourself. Second, voice: Verke's voice sessions run up to twenty minutes, and a written summary posts back into the chat so you can pick up in text the next day. Therabot's trial used text-only interactions; Verke's voice modality exists because phone-call-shaped sessions are how a lot of coaching actually feels.

In the Stockholm University trial, participants often described being able to reach their coach in the moment they needed it, getting concrete tools to try, and coming to understand their own patterns. They also described finding it easier to open up when there was no one to judge them. Therabot's NEJM AI outcomes are one published clinical signal in this space; the Stockholm trial, published in Internet Interventions in 2026, is Verke's parallel evidence trajectory.

• AI therapy vs human therapy: a deeper look
• Is AI therapy safe?
• Find your AI coach (quiz)
• Types of AI therapy