Amplitude Uses OpenAI Models to Ship Improvements in Days

Amplitude published details about how it uses OpenAI models to power its proactive product agent, Wave, which helps teams identify product opportunities, coordinate implementation with coding agents like Codex, and measure outcomes.

Wave is built primarily on OpenAI models. GPT-6.1 Sol supports opportunity management and code investigations, GPT-6 Luna handles frequent low-cost tasks, and GPT-Image-2.5 Flare generates product mockups.

According to Amplitude, Wave helps teams ship product improvements in 1–2 days. In an internal test on the Amplitude homepage, Wave proposed personalizing suggestion cards, resulting in a 160% lift in agent engagement compared to a control group after one day.

During evaluations for Wave’s Opportunity Investigator, GPT-6.1 Sol scored 4.06 out of 5 overall with a median investigation time of 54 minutes. In evaluations for the Data Science Agent, GPT-6.1 Sol scored 67.5 out of 100 with a mean model cost of $1.28 per dispatch. For wireframe evaluations, GPT-Image-2.5 Flare scored 80.3 out of 100 for solution wireframes.