China Dropped a 2.8 Trillion Parameter Open Model
Hours Before Google’s Biggest Launch of 2026
July 17 was the single biggest AI day of 2026 so far. China’s Moonshot AI dropped the largest open model in history hours before Google’s rebuilt flagship was due to launch, while Xi Jinping personally opened the World AI Conference proposing a new global governance body. TSMC printed a record quarter, and Mira Murati shipped her first model since leaving OpenAI. Here’s everything that happened.
Kimi K3 Drops as the Largest Open Model Ever — 2.8 Trillion Parameters
Moonshot AI launched Kimi K3 late on July 16, a sparse Mixture-of-Experts model with roughly 2.8 trillion total parameters, a 1-million-token context window, native vision, and always-on reasoning. It shipped in two variants — K3 Max for chat and agent tasks and K3 Swarm Max for large-scale parallel processing — at $3 input and $15 output per million tokens. Open weights are promised by July 27, which would make it the largest open-weight release in AI history. Early independent testing places it around the Opus 4.8 and GPT-5.5 tier, competitive with top closed coding models. The timing was surgical — Moonshot dropped a frontier-class model hours before Google’s biggest launch of the year.
Gemini 3.5 Pro Targets July 17 After Google Scrapped and Rebuilt the Model
Google’s Gemini 3.5 Pro was widely expected to launch on July 17, arriving weeks late after Google scrapped the original base model entirely and restarted pretraining from scratch. Engineers reportedly found structural failures in recursive tool-calling and SVG generation. Leaked specs point to a 2-million-token context window, a Deep Think reasoning mode on the $250/month Ultra tier, and API pricing near $1.25 input and $10 output per million tokens — but Google has never officially confirmed the date, specs, or pricing. The rebuild signals Google refused to ship a flawed flagship, but it also means the launch carries the pressure of a do-over, landing after both GPT-5.6 Sol and Grok 4.5 have already reset the competitive field.
Xi Jinping Keynotes the World AI Conference — Proposes Global AI Governance Body
Chinese President Xi Jinping delivered the opening keynote of the 2026 World AI Conference in Shanghai on July 17 — his first appearance at the event since it launched in 2018. The conference runs July 17–20 with more than 140 forums and 1,100+ exhibitors, and this year it doubles as a High-Level Meeting on Global AI Governance. The headline move: China’s proposed World AI Cooperation Organization, an international governance body Beijing wants headquartered in Shanghai. In practical terms, China is volunteering to write the global AI rulebook at the moment Washington has no equivalent proposal on the table. The Kimi K3 launch the night before Xi’s speech was not a coincidence of calendars.
TSMC Q2 Profit Surges 77% — Raises US Investment to $265 Billion
TSMC reported second-quarter net profit of roughly $22 billion, up 77% year over year, on revenue of $40.2 billion (up 34%), beating analyst expectations on sustained AI chip demand. The company raised its 2026 capital spending forecast to $60–$64 billion, guided full-year revenue growth above 40%, and announced an additional $100 billion for its Arizona operations — lifting total planned US investment to $265 billion with up to four more fabs focused on advanced 2-nanometer manufacturing. The pattern of 2026 continues: model companies cut prices while the hardware layer prints records. Every price war requires more silicon, and TSMC collects on every side.
Mira Murati’s Thinking Machines Releases Inkling — Its First Open-Weight Model
Thinking Machines, the startup founded by former OpenAI CTO Mira Murati, released Inkling, an open-weight model that anyone can download, run locally, and fine-tune. It’s the company’s first major public release since its heavily funded founding, and the format is the message: the executive who helped build the most famous closed models in the world chose open weights for her independent debut. Inkling joins a 2026 open-model surge that now looks like a stampede — DeepSeek, Qwen, GLM, Kimi, and Bonsai 27B all holding open-weight positions, with an American marquee founder now on the same side of the ledger.
That’s today’s wrap. The split-screen of July 17 is the whole year in one image — the West’s biggest model launch, the East’s biggest AI policy speech, and a Chinese open model dropped between them, all within 24 hours. Follow NeuralPaws daily for the next AI news drop.