Pinterest CEO Highlights Open‑Source AI for Cost‑Effective Visual Services
Key Points
- Pinterest is testing open‑source AI models against proprietary alternatives.
- Open‑source models show comparable performance with dramatically lower costs.
- Cost reductions are described as orders‑of‑magnitude versus leading models.
- The company plans to expand open‑source AI across recommendation, search, and ad targeting.
- Pinterest Assistant, an AI conversational tool, has been launched for personalized advice.
- Agentic commerce is being explored, but direct AI‑driven purchases will be monitored.
- A partnership with Amazon enables push‑button buying for users.
- AI‑curated personalized boards combine human expertise with machine learning.
- Stock fell over 21% after a revenue forecast miss.
On a recent earnings call, Pinterest chief executive Bill Ready explained how the company is turning to open‑source artificial‑intelligence models to slash costs while preserving performance across its visual‑AI features, including personalized recommendations, multimodal search, ad targeting and the new AI‑powered Pinterest Assistant. Ready said tests show open‑source models delivering comparable results at a fraction of the price of proprietary alternatives, offering what he described as orders‑of‑magnitude cost reductions. The company plans to expand the use of these models, while also exploring agentic commerce options and AI‑curated boards. The announcement came as Pinterest’s stock fell more than 21% after a revenue forecast miss.
Open‑Source AI Strategy
Pinterest’s chief executive Bill Ready told investors that the company is actively evaluating leading off‑the‑shelf AI models against open‑source alternatives. In early testing, the open‑source models delivered "tremendous performance" for Pinterest’s visual‑AI use cases while dramatically lowering per‑token costs. Ready characterized the cost advantage as an "orders of magnitude reduction" compared with leading proprietary models. By fine‑tuning these open‑source solutions, Pinterest aims to support its core visual‑AI functions—including personalized recommendation engines, multimodal search that blends text and images, and ad‑targeting algorithms—without a significant increase in its cost structure.
The company already runs its own proprietary models, but Ready emphasized that the new open‑source approach will complement existing infrastructure and be deployed across a variety of product features. He noted that this strategy aligns with Pinterest’s broader goal of delivering value to users while keeping operating expenses in check. The emphasis on cost‑effective AI comes as the firm navigates a challenging revenue outlook, which contributed to a more than 21% drop in its stock price following the earnings release.
Future AI Initiatives
Beyond cost savings, Pinterest is expanding its AI capabilities into new consumer experiences. The firm recently launched the AI‑powered Pinterest Assistant, a conversational companion that can answer user queries, offer recommendations, and understand individual tastes based on boards, saved pins, and comparisons with similar users. Ready indicated that the assistant is part of a broader push toward “agentic commerce,” where AI systems could act autonomously on behalf of shoppers. However, he cautioned that Pinterest will monitor user demand before allowing the AI to execute purchases directly, noting that the platform already supports a “push‑button type buying” experience through its partnership with Amazon.
In addition, Pinterest is rolling out AI‑curated personalized boards that blend expert human curation with machine‑learning insights. These boards aim to enhance discovery and keep users engaged by surfacing content that matches their interests. Ready expressed confidence that these initiatives, combined with the cost efficiencies of open‑source AI, position Pinterest to strengthen its role in the visual discovery and shopping journey.