Chinese AI

Daily Chinese AI updates

• Paris-based Mistral releases "Le Chonk," a strong coding model that still trails China overallThe Decoder · 29 Oct 2026
• Mistral’s Large 4 open weights arrive with a 1-point coding lead over GLM-5.3: "There’s nothing you can’t do with this model that you can do with competitors."The New Stack · 29 Oct 2026
• Bombshell Report Claims White House Did Nothing After Chinese Spy Targeted Nvidia CEO Jensen HuangMediaite · 29 Oct 2026
• China’s Iluvatar CoreX says new GPU matches Nvidia’s H200 as chip gap closesSouth China Morning Post · 29 Oct 2026
• Half of healthcare’s $1 trillion admin spending is wasted, and Utah is tackling it with AIUtah Business · 29 Oct 2026
• Rival.io Launches GraphX, Running DeepSeek V4 Flash on Just Eight NVIDIA L4 GPUsYahoo Finance · 29 Oct 2026
• Mistral Says Its AI Model ‘Le Chonk’ Has Leapfrogged China in the Open-Weight RaceWired AI · 29 Oct 2026
• What Latino Voters Can Teach Washington About AI GovernanceTech Policy Press · 29 Oct 2026
• Let China copy American AIThe Economist · 29 Oct 2026
• Chinese spies targeted Nvidia CEO Jensen Huang, according to The New York TimesThe Decoder · 29 Oct 2026
• Regulate AI, but don’t hand the 21st century to ChinaThe Hill · 29 Oct 2026
• OpenAI s Astra becomes the main target for Chinese AI labs after Anthropic tightens its defensesThe Decoder · 29 Oct 2026
• New Policy Memo: An American Answer to China’s Open AI ModelsManhattan Institute · 29 Oct 2026
• Guinness Brings its Ad Characters to China in AI Campaign FilmBranding in Asia · 29 Oct 2026
• China's Rokae Robotics Announces Delivery of its 10,000th Humanoid Force-Controlled ArmAI Insider · 29 Oct 2026
• Bio Essence Unveils Practitioner-Controlled Knowledge Vault, Tackling AI's Trust Barrier in HealthcareYahoo Finance · 29 Oct 2026
• IMF proposes Min Liao to succeed Bo Li as deputy chiefThe Financial Express · 29 Oct 2026
• IMF proposes China’s Min Liao as deputy managing director, Bo Li to exit in NovemberThe Economic Times · 29 Oct 2026
• Nvidia Chief, Jensen Huang, Is Seen as Target of Chinese Spy PlotNYT Technology · 29 Oct 2026
• Nvidia Chief Seen as Target of Chinese Spy PlotThe New York Times · 29 Oct 2026
• India AI Race: Why Private Equity and Family Offices Still See Promise Despite a Funding Gap With the US and ChinaUnivest · 29 Oct 2026
• DeepSeek and Moonshot AI are reportedly using GPT-6 Astra for trainingTechzine Global · 29 Oct 2026
• Chinese AI Models Under Scrutiny as Cyberattacks Target Indian Financial FirmsThe 420.in · 29 Oct 2026
• Chinese Startup DiffuSpace Open-Sources 'Clover,' a 35-Billion-Parameter Diffusion Language ModelBigGo Finance · 29 Oct 2026
• Chinese AI model completes complex cyberattacks in tests, raises concernsBusiness Standard · 29 Oct 2026
• Kimi vs DeepSeek: Which Chinese AI Should You Use?Fello AI · 29 Oct 2026
• DeepSeek V5 identifiers surface in official Harness code repositoryTechNode · 29 Oct 2026
• China AI Export Controls: 40 Technologies Listed [2026]https://tech-insider.org/ · 29 Oct 2026
• China calls for greater regulation on AI at global forumModern Ghana · 29 Oct 2026
• DeepSeek GPT-6 Astra: 2 Models Named in Report [2026]shattered.io · 29 Oct 2026

Top Chinese open models

1. Qwen/Qwen3.8-27B♥ 18665
2. Qwen/Qwen3.8-Flash-Next♥ 6635
3. MiniMaxAI/MiniMax-H3♥ 6433
4. baidu/Unlimited-OCR♥ 4888
5. deepseek-ai/DeepSeek-V4.1-Flash♥ 4724
6. deepseek-ai/DeepSeek-V4.1-Flash-1017♥ 3708
7. Qwen/Qwen-Image-2.1♥ 3304
8. openbmb/MiniCPM-V-4.7-35B-A3B♥ 1071
9. Qwen/Qwen-Image-2.1-Turbo♥ 1034
10. MiniMaxAI/MiniMax-M3.1-Flash♥ 479
11. tencent/Hy-SimulMT-30B-A3B-FP8♥ 456
12. openbmb/MiniCPM-V-4.7-1B♥ 394
13. baidu/Unlimited-OCR-2♥ 369
14. tencent/Hy-Memory-30B-A3B♥ 313
15. openbmb/MiniCPM-OCR-1.0♥ 306
16. openbmb/MiniCPM-V-4.7-35B-A3B-gguf♥ 186
17. stepfun-ai/Step-5-Preview♥ 122
18. tencent/Hy-Nomas♥ 110
19. Qwen3-0.6B30,452,693
20. Qwen3-4B12,585,505
21. Qwen3-8B9,739,109
22. Qwen3-Embedding-0.6B8,598,074
23. Qwen3.5-9B8,489,640
24. Qwen3.5-4B7,946,136
25. Qwen3-VL-8B-Instruct7,390,307
26. Qwen2.5-7B-Instruct7,357,696
27. Qwen3.8-27B7,221,327
28. Qwen3-14B6,913,178
29. GLM-5.3-Flash6,515,511
30. Qwen2.5-0.5B-Instruct6,276,110
31. Qwen2.5-1.5B-Instruct6,018,399
32. Qwen2.5-VL-7B-Instruct5,132,099
33. DeepSeek-V3.24,844,451
34. Qwen3.8-27B-FP84,470,739
35. DeepSeek-V4-Flash-07314,438,887
36. Qwen2.5-3B-Instruct4,299,162
37. Qwen3-1.7B4,034,338
38. Qwen3.6-35B-A3B-FP84,021,372
39. MiniMax-H33,765,471
40. Qwen3.6-35B-A3B3,667,555
41. Qwen3-4B-Instruct-25073,293,032
42. Qwen3.5-2B3,243,646
43. Qwen3-32B3,235,453
44. Qwen3-VL-4B-Instruct3,208,816
45. Qwen3-TTS-12Hz-1.7B-Base3,162,646
46. Qwen3-14B-AWQ2,894,022
47. Qwen3.5-0.8B2,820,039
48. Qwen3-VL-2B-Instruct2,385,218