← Back to Model Beat
Models·5d ago·all news from September 18, 2026

Qwen3.8-Omni-Flash undercuts Google's Gemini Flash pricing while matching its multimodal benchmarks

Zhipu has released GLM-5.3-FlashX, a model optimized to achieve speeds nearing 200 tokens per second. This performance is supported by an infrastructure of approximately 100,000 domestic Chinese accelerators.

ModelsQwen3.8 Omni FlashGLM 5.3 FlashXGLM-5.3-Flash

Covered by 8 sources · 12 articles

Related stories

ModelsGLM-5.3-Flash matches top models at a fraction of the cost, and runs without NvidiaAug 26 · 11 sourcesModelsGLM-5.3 tops the open-model rankings and undercuts rivals on price, but its release is delayedAug 17 · 6 sourcesOpen SourceChina's Zhipu Says Open-Source GLM-5.3 Outperforms Anthropic's Restricted Model in Vulnerability DetectionAug 14 · 7 sourcesModelsAlibaba's Qwen3.8-Omni-Flash Slashes Audio Pricing 98% and Drops Open WeightsSep 20