Reportedly: Zhipu AI Gray-Releases GLM-5.3-Highspeed in ZCode, Users Measure About 490 Tokens/s
Model: GLM-5.3, GLM-5 · availability, license and releases
Reported by Pandaily · not yet confirmed by the company or a second independent outlet. We update this page when it is.
Zhipu AI is quietly testing GLM-5.3-Highspeed with randomly selected ZCode subscribers, Pandaily reported. One user measured about 490 tokens/s; Zhipu has not published an official speed figure.
China context
- Original name
- 智谱
- Outside China
- Not stated in the sources yet
- Claims
- Company-reported; not yet independently evaluated
- For builders
- If the speed is confirmed, developers outside China may see competitive pressure on inference pricing and latency from Chinese providers, but no API availability outside China is indicated.
- For investors
- Investors should monitor whether Zhipu AI officially confirms the speed and expands access, as this could affect the competitive landscape for high-throughput inference.
Zhipu AI has begun a gray release of GLM-5.3-Highspeed within its ZCode platform, according to Pandaily. The model is being tested with randomly selected subscribers, and one user reported a speed of approximately 490 tokens per second. Zhipu has not officially confirmed this figure or announced a public release date.
The reported 490 tokens/s is a user measurement, not an official benchmark. No independent evaluator has confirmed this speed. The model's architecture, parameter count, and context window are not disclosed in the evidence.
If confirmed, GLM-5.3-Highspeed's reported speed could pressure competitors to match or exceed it in high-throughput inference, potentially lowering costs for developers using ZCode. However, without official benchmarks or independent verification, the competitive impact remains unproven.
For developers, a high-speed model could reduce latency and infrastructure costs for real-time applications. For investors, confirmation of the speed and adoption metrics would signal Zhipu's competitiveness in the inference market.
Watch for Zhipu AI to publish official speed benchmarks or announce a wider rollout of GLM-5.3-Highspeed. Independent evaluations comparing it to other high-speed models would clarify its market position.