China's Zhipu AI claims to have developed the GLM-5.2 model, matching Anthropic's Mythos in cybersecurity tasks. However, it still falls short compared to models from Anthropic and OpenAI in other tasks.
| # | Наименование новости | Тональность | Информативность | Дата публикации |
|---|---|---|---|---|
| 1 | China’s Z.ai unveils GLM-5.3, claims chart-leading scores | 0 | 16.76 | 14-08-2026 |
| 2 | Новый AI игрок на поле долгосрочных задач | 0 | 9 | 02-07-2026 |
| 3 | Z.ai составит конкуренцию Anthropic и OpenAI в программировании с помощью новой модели ИИ GLM-5.3 | 0 | 13.45 | 14-08-2026 |
| 4 | Китайский конкурент OpenAI и Anthropic запустил дешёвый аналог Claude Code и Codex | 0 | 5 | 07-07-2026 |
| 5 | Open-weight AI models are catching up to the frontier. The safety gap remains. | 0 | 11.45 | 04-08-2026 |
| 6 | What is Claude Fable 5? 7 things to know about Anthropic's first Mythos model | 0 | 5 | 10-06-2026 |
| 7 | Chinese AI developer Z.AI may enter Russian market | 0 | 9.22 | 27-07-2026 |
| 8 | GLM-5.3 is here with advanced cyber capabilities — and reportedly already found a 'serious vulnerability' in Cursor | 0 | 6.87 | 14-08-2026 |
| 9 | AI's double-edged (cyber) sword | 0 | 7 | 21-05-2026 |
| 10 | OpenAI ответила на Claude Mythos 5. Вышла GPT-5.6 сразу в трех версиях, которые лучше всего подходят для кибербезопасности и биологии | 5 | 7 | 26-06-2026 |