Testing Qwen3.8-Max: Open Models Are Catching Up with Closed Frontier
dair_ai · x · 2026-08-04
AI news account DAIR.AI shared hands-on impressions of Alibaba's Qwen3.8-Max. The reviewer noted that trying the model on Hermes Agent forces you to doubt how much open frontier models have caught up with closed ones.
The reviewer praised the new open models, stating they are "insanely good" and demonstrating extremely high capability levels in agentic tasks.
More from Models
- A single image puts multiple AI models in one frame — wavefnx · 2026-08-04
- Databricks says Kimi K3 now runs at 239 tokens per second on its serving stack — Yuchenj_UW · 2026-08-04
- Claude Max users debate whether to add Codex Pro, GLM 5.2 or Kimi K3 for coding — tinker_20 · 2026-08-04
- Verifying GPT Pro: Zero Math Errors, But Disastrous Exposition — josh_wills · 2026-08-04
- Ant Group Engineer's Long Post: The Four Very Different Bets of Chinese AI Labs — AcanthisittaOk1699 · 2026-08-04
- OpenAI's Upcoming 'Astra' Model to Focus on Multi-Agent Collaboration — thesaraharminta · 2026-08-04