In its latest earnings call, Pinterest unveiled a major strategic shift, embracing open source AI models to power its growing suite of visual and recommendation features. According to CEO Bill Ready, the move could deliver “orders of magnitude” reductions in AI operating costs while maintaining performance comparable to top proprietary models.
The decision marks a pivotal moment for Pinterest as it balances AI innovation with cost efficiency, leveraging community-driven technology to scale visual search, shopping personalization, and its new AI assistant.
Open Source as a Cost-Cutting Advantage
During the Q3 2025 earnings call on November 4, Ready highlighted the company’s success in testing open source models against leading proprietary systems. “We’re getting tremendous performance from open-source models, specifically for Pinterest use cases on visual AI,” he said, noting early trials achieved massive cost savings with “fine-tuned versions matching premium alternatives.”
CFO Julia Donnelly reinforced the point, calling open source “a key part of our cost-efficient AI strategy” to sustain long-term R&D investments. Both executives emphasized that Pinterest’s side-by-side testing ensures open source can meet commercial-grade standards without compromising on quality or reliability.
Record Growth Despite Market Pressures
Pinterest’s financial performance underlines the importance of its AI-driven cost control strategy. The company reported $1.049 billion in Q3 revenue, up 17% year-over-year, with global monthly active users reaching 600 million, a record high. Adjusted EBITDA rose 24% to $306 million, reflecting healthy operational leverage.
However, investor concerns over a weaker Q4 outlook and external factors like tariffs on home furnishings led to a 21% stock drop following the announcement. Still, Pinterest remains optimistic about its long-term growth, with open source adoption positioned as a core pillar of its AI scaling roadmap.
Also Read: Pinterest Predicts 2025: Emerging Trends Overview
Visual AI at the Core of Pinterest’s Future
Pinterest’s AI strategy is deeply embedded in its platform. From multimodal search that blends text and images to the Pinterest Assistant, an AI-powered conversational shopping guide, the company is building what Ready describes as “agentic commerce”, where AI proactively assists users through the discovery and buying journey.
By integrating open source models into these systems, Pinterest aims to accelerate personalization while keeping AI training and inference costs manageable. Industry insiders suggest the company is fine-tuning visual foundation models, possibly derived from frameworks like Llama or Mistral, for use in recommendations and visual search.
Industry and Market Reaction
The tech community has largely welcomed Pinterest’s open source shift. Hugging Face CEO Clem Delangue praised the company’s approach on LinkedIn, calling it a “pragmatic step toward scalable visual AI.” Others echoed the sentiment, noting that open source provides flexibility and transparency, even if it demands stronger internal governance and data security.
Analysts see this as part of a broader trend, major tech firms seeking alternatives to expensive proprietary AI systems amid rising compute costs and uncertain macroeconomic conditions.
Outlook: Balancing Innovation and Efficiency
Pinterest plans to expand its use of open source AI in 2026 to sustain 30–34% EBITDA margins over the next few years. With strong user engagement and advertiser adoption, the company is betting that efficient AI infrastructure can fuel growth without spiraling costs.
Still, challenges remain — from privacy regulations to integration risks. Yet if executed well, Pinterest’s pivot could make it a model for cost-effective AI transformation in consumer tech, proving that open source innovation can rival even the biggest proprietary platforms.
Pinterest Expands AI Strategy with Focus on Agentic Shopping and Personalized Experiences
CEO Bill Ready further emphasized that Pinterest’s evolving use of open-source AI is part of a broader effort to enhance user value and ensure sustainable monetization.
“We feel outstanding about the value that we’re bringing to the user there, our ability to align that with monetization, and our ability to control those costs and deliver that effectively,” Ready said during the Q3 2025 earnings call.
Beyond cost efficiency, Ready discussed how Pinterest is exploring “agentic shopping” — where AI not only recommends products but actively assists users throughout their purchase journey. He highlighted that Pinterest already enables “push-button type buying” through its partnership with Amazon but noted that the company will observe whether users actually want AI to go as far as making purchases autonomously.
“We’ll see if people want the AI to ‘push the button for them,’” Ready added, underscoring the company’s cautious and user-centered approach to automation.
In the meantime, Pinterest is positioning itself as a guide rather than a mere marketplace. Ready described the company’s biggest differentiator as its ability to lead users through inspiration-driven shopping, helping them refine their preferences, discover new products, and make informed choices rather than executing quick transactions.
This experience is being deepened with Pinterest Assistant, the company’s new AI companion that users can chat with for personalized advice and recommendations. The assistant leverages Pinterest’s deep understanding of each user’s activity, including their boards, collages, and saved Pins- and compares those behaviors to similar taste profiles across the platform.
Pinterest is also introducing AI-curated, personalized boards that blend expert human curation with machine intelligence to deliver more relevant, context-aware collections. According to Ready, this hybrid approach ensures that the platform maintains its signature human touch while scaling personalization through AI.
By combining open-source model efficiency, agentic commerce, and conversational interfaces, Pinterest aims to solidify its position as the leading visual discovery engine for lifestyle and shopping, all while managing costs amid a competitive AI landscape.
