Zhipu GLM-5.3-Flash: Matches Opus 4.8 at 1/40 the Cost, Powered by Domestic Chips
vista8 · x · 2026-08-27
Zhipu AI revealed that the viral "Ox-Alpha" model is officially named GLM-5.3-Flash. It features 320B total parameters with only 18B active, marking the first natively multimodal model in the GLM-5 series.
- Performance: Scored 57 on the Artificial Analysis Index, matching Opus 4.8 and surpassing GLM-5.2.
- Pricing: Offered at a limited-time discount of 1/20th of GLM-5.3's price and 1/40th of Opus 4.8's.
- Infrastructure: The daily trillion-scale token throughput is supported entirely by domestic Chinese chips.
The author includes a test case demonstrating the model's ability to OCR an image-only PDF and generate a multilingual corporate website, highlighting its multimodal understanding and generation quality.
Related event: Zhipu Open-Sources GLM-5.3-Flash: 320B MoE Multimodal at 1/10th the Cost(48 posts)→
More from Models
- Unverified: internal snapshot labeled gpt-6-astra-api-priv-ev3 spotted on model tracker — lyraxana · 2026-10-16
- NVIDIA releases Nemotron 3.5 Bolt 74B MoE that fits in 24GB VRAM at 2-bit — NerdyRodent · 2026-10-16
- Muse Code 1.4.8 cuts token usage; all user quotas reset immediately — alexandr_wang · 2026-10-16
- Alibaba and Baidu to power key parts of Apple's AI services in China — nordicinst · 2026-10-16
- Cornell and Amazon revisit the 16x pretraining compute gap: diffusion LMs need only 3-8x — LucaAmb · 2026-10-16
- Image Decision Benchmark JevImageBench hits v0.3.0 with reshuffled Top 5 — airesearch12 · 2026-10-16