Zhipu GLM 5.3 Flash Beats Rivals on OfficeQA Benchmarks
jefrankle · x · 2026-08-28
Zhipu released GLM 5.3 Flash, a frontier model in the efficient tier. It outperforms DeepSeek V4 Flash 0731 and GPT 5.6 Luna on OfficeQA Pro and Pro V2 benchmarks, pushing the quality-cost pareto frontier. Its vision capabilities approach Gemini 3.6 Flash on enterprise document parsing tasks. Available on Databricks.
More from Models
- Qwen3.8-Flash reportedly costs 1/9th to train vs Qwen3.7-Plus — VraserX · 2026-08-28
- Qwen3.8-Next paper: matches 397B predecessor with 1/9 the training FLOPs — NielsRogge · 2026-08-28
- DeepSeek retakes the throne on day one of GLM going paid; muse spark a sleeper hit — drdanielbender · 2026-08-28
- Qwen3.8-Flash lands in OpenCode Go: 125B/6B, 1M context, multimodal — Alibaba_Qwen · 2026-08-28
- Tencent Hunyuan Releases Hy4 Preview: 770B Params, 1M Context — TencentHunyuan · 2026-08-28
- Anthropic's 'Model Welfare' Is Making Claude Worse as an Assistant — Nouni2 · 2026-08-28