Moonshot AI Kimi K3 Disrupts US Markets as Tech Firms Adopt Chinese AI Models

A fresh wave of affordable artificial intelligence software from Chinese startups is altering how developers work in the United States. Software engineers and major crypto platforms are shifting toward overseas models to cut costs.

Mozilla and Coinbase Shift Toward Chinese AI Models

In San Francisco, Mozilla Chief Technology Officer Rafi Krikorian adopted Moonshot AI’s Kimi K3 model for everyday tasks just days after its July release. He noted that the software feels faster and more responsive, comparing it directly to Anthropic’s Claude Fabius, a high-end chatbot based in the same city that carries a higher cost.

View this post on Instagram about moonshot kimi disrupts markets, ديب سيك الصين

Before adopting Kimi K3, Krikorian relied on GLM 5.2 from Zhipu AI—also known as Chipu—to manage routine tasks like calendars, documents, and emails. They are not alone in this shift, as a growing number of Americans embrace these systems to benefit from lower costs and improved efficiency. U.S. companies, including the cryptocurrency trading platform Coinbase, are moving toward Chinese models to help reduce operational expenses.

Silicon Valley Startups Build Cheap AI Models to Rival Chinese Systems

Kimi K3 Disrupts US Benchmarks and Wall Street

Launched on a Thursday in July, Kimi K3 quickly topped the widely watched Arena leaderboard for AI programming tools within hours, marking a milestone for a Chinese model. Investors in Silicon Valley and Wall Street, alongside White House officials, are growing anxious that if China can build artificial intelligence models matching the quality of American counterparts, firms like OpenAI and Anthropic may struggle to maintain high pricing structures for their products.

Anastasius Angelopoulos, director of the Arena platform, addressed these market impacts during an appearance on the TITV podcast. Angelopoulos warned that the release could prompt a broader reassessment across financial markets, noting that companies might prefer free Chinese software that they can customize and run on their own devices over paid American software that requires data-sharing with external providers. He suggested the model raises questions about the dominance of U.S. closed-source systems and could trigger a radical re-evaluation in financial markets.

Global Markets Rally as US and Iran Pause Hostilities and Oil Prices Plunge

Annahar, Technology Desk

The Shadow of DeepSeek and Washington Policy Debates

Industry observers trace the current market unease back to early 2025, when DeepSeek stunned global markets by releasing a powerful AI model at a fraction of standard costs, briefly wiping hundreds of billions of dollars from U.S. tech valuations. Like most Chinese offerings, Kimi K3 delivers lower costs alongside code that lets programmers make adjustments.

Zhipu AI debuted its high-performing GLM 5.2 model in June, while Moonshot AI launched Kimi K3 in July. In the same month, Alibaba demonstrated its Qwen 3.8 Max model, and DeepSeek rolled out trial versions of its V4 model in April to challenge OpenAI’s GPT and Google’s Gemini systems.

Political friction is mounting alongside these technical gains. Washington-led restrictions prevent China from accessing advanced global technologies like sophisticated AI chips, and Treasury Secretary Scott Bessent has warned of potential additional sanctions to protect American intellectual property. Meanwhile, the administration of President Donald Trump accused Moonshot AI of using “secretive” methods—though not necessarily illegal ones—to build Kimi K3 by basing it on Anthropic’s Fabius model. Some American politicians and AI companies, including Anthropic, have accused Chinese startups of illicitly using distillation technology to extract model capabilities, accusations that Beijing has rejected as “baseless.”

Arcee AI Builds Affordable Open Model to Counter Cheap Chinese AI

David Sachs, AI Advisor to the White House

David Sachs, a venture investor advising the White House on artificial intelligence and a vocal opponent of regulatory restrictions on the sector, argued via social media that legislative red tape damages domestic competitiveness. He criticized politicians for slowing data center construction, imposing state-level regulations, and pushing for a federal agency to approve powerful AI models before release, stating, This is how you lose the AI race.