Taobao's RecGPT-V3 Recommendation System Tech Report
Bowen Zheng · hf · 2026-07-20
Large language models are shifting recommendation systems from "matching historical behavior patterns" to "reasoning behind intent." To address computational waste, information bottlenecks, and inference latency challenges faced by previous versions at scale, the Alibaba Taobao team launched RecGPT-V3.
Architecture Upgrades & Business Gains:
- Memory Hub: Maintains structured, continuously evolving user memories, distilling long-term behaviors into condensed units, slashing user modeling compute by 55.8%.
- Hybrid Modality Reasoning: Combines natural language and semantic IDs (SIDs) for joint reasoning, bridging open-world knowledge with specific items.
- Latent Intent Reasoning: Internalizes lengthy chains of thought into compact latent tokens, cutting output token costs by 200x while remaining decodable into human-readable explanations.
- Online Performance: Deployed in Taobao's "Guess What You Like" feed, achieving CTR +1.00% and GMV +3.97% in A/B testing, alongside a 52.4% reduction in end-to-end serving resource consumption.
Related event: Alibaba Taobao Unveils RecGPT-V3 Recommendation LLM(2 posts)→
More from Apps
- Google Docs and Drive add native Markdown support for human-AI collaboration — thisiskp_ · 2026-10-27
- Instinct launches Link balance to let agents spend within set budgets — jeff_weinstein · 2026-10-27
- Open Dungeon Master: open-source AI Dungeon Master app hits Play Store, runs 5e on a 16GB GPU — lebbi · 2026-10-27
- Show HN: Wraps, open-source Resend-style email API running on your own AWS SES — stewartjarod · 2026-10-27
- Sentry CEO builds AI-generated entity pages for whisky brands in side project Peated — zeeg · 2026-10-27
- Grok Impressions: Delegating Forms, Shopping and Payments to Grok Bot — XFreeze · 2026-10-27