Moonshot pauses Kimi K3 signups amid GPU shortage

2026-07-20 · Source: Dataconomy · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Emerging Technologies & Innovation · Depth: Intermediate, quick

Summary

Moonshot AI temporarily suspended new consumer signups for its Kimi K3 model on July 19, just three days after its July 16 launch, due to overwhelming demand straining its computing infrastructure. The Kimi K3 is a 2.8 trillion-parameter mixture-of-experts model featuring a one-million-token context window and native vision capabilities, making it the largest open-weight system approaching the 3-trillion-parameter mark, surpassing DeepSeek V4's 1.6 trillion parameters. User request volume in the first 48 hours exceeded projections, nearing Moonshot's GPU capacity. In response, the company will restructure its membership into "Kimi Membership" for general use and "Kimi Code Membership" for programming, aiming for better resource allocation. Kimi K3 achieved a score of 57 on Artificial Analysis's Intelligence Index, ranking closely behind Anthropic's Claude Fable 5 (60) and OpenAI's GPT-5.6 Sol (59), and outperformed rivals in front-end coding and long-context tasks. Its API pricing is \$3 per million input tokens and \$15 per million output tokens, with full model weights expected to be open-sourced by July 27.

Key takeaway

For AI Product Managers launching new large language models, you must anticipate extreme demand surges and plan for scalable compute infrastructure. Your launch strategy should include mechanisms like tiered access or phased rollouts to protect existing user experience. Consider open-sourcing model weights post-launch to distribute compute load and foster community development, mitigating infrastructure strain.

Key insights

Rapid user adoption of large AI models can quickly overwhelm even substantial computing infrastructure.

Principles

• Resource allocation benefits from specialized product tiers.
• Open-sourcing models can decentralize compute demand.

In practice

• Monitor compute capacity closely post-launch for new models.
• Consider tiered access or open-source releases to manage demand.

Topics

• Large Language Models
• GPU Shortage
• AI Infrastructure
• Kimi K3
• Mixture-of-Experts
• Open-Source AI

Best for: CTO, VP of Engineering/Data, AI Architect, Tech Journalist, Director of AI/ML, AI Product Manager

Related on AIssential

• Moonshot AI launches Kimi K3 — Kimi K3's release intensifies open model competition, potentially narrowing the performance gap with proprietary LLMs.
• Why Google is Limiting Meta’s Gemini AI Use as Demand Soars — Global AI expansion faces severe compute infrastructure bottlenecks, impacting even major tech companies' growth and operations.
• Moonshot pauses new Kimi K3 subscriptions after GPU demand maxes out in 48 hours — Intense demand for advanced AI models rapidly exhausts GPU capacity, challenging service providers.
• Just like Deepseek, China's Kimi K3 is forcing Western AI labs to question their compute advantage — Chinese AI labs are challenging Western compute dominance through innovation and efficient model development.
• Is the World’s Largest Open-Source AI Model Worth the Hype? — A model's "open-weight" status does not guarantee practical local deployment for massive, trillion-parameter systems.
• Kimi K3 threatens AI business models — The rise of open-weight models like Kimi K3 challenges frontier AI labs' business models and forces a dilemma between open innovation and proprietary secrecy.

See Counsel's argued verdicts on the open AI decisions leaders are weighing →

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Dataconomy.