Zhipu AI has reportedly released a new model, GLM-5.3-Highspeed, within its Zcode platform, with some users experiencing speeds of up to 490 tokens per second. This represents a significant speed increase compared to previous models like GLM-5.3-FlashX, which reached 200 tokens/s, and GLM-5.3-Flash at 60 tokens/s. The new model's name suggests a focus on speed, and it may be related to the Omen Alpha model found on OpenCode, though its exact identity and capabilities are still under discussion among users. AI
IMPACT Potentially sets a new benchmark for LLM inference speed, impacting real-time applications and long-context processing.
RANK_REASON New model release from a frontier lab (Zhipu AI) with performance metrics. [lever_c_demoted from frontier_release: ic=1 ai=1.0]
- Artificial Analysis
- GLM-5.1-highspeed
- GLM-5.3-FlashX
- GLM-5.3-Highspeed
- Omen Alpha
- OpenCode
- Zcode
- Zhipu AI
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
