Pyyan / Compare / GLM-5.3-Flash vs Atria Dawn Preview
GLM-5.3-Flash vs Atria Dawn Preview
2 of 5Open-Weight Models · verified 16 Sept 2026
GLM-5.3-FlashZhipu AIcurrentSAtria Dawn PreviewShanghai AI Laboratorycurrent
| Specification | GLM-5.3-Flash | Atria Dawn Preview |
|---|---|---|
| Summary | MIT licensed, a million tokens of context, fifteen cents a million in. | 744B agentic weights under the MIT licence, published without a launch. |
| Context | 1M | 256K |
| Input ($/Mtok) | $0.15 | Not published |
| Output ($/Mtok) | $0.5 | Not published |
| Licence | MIT | MIT |
| Max output | 32K | Not published |
| Parameters | 320B total, 18B active | 744B total |
| Released | 26 Aug 2026 | 14 Sep 2026 |
| Model ID | zai-org/GLM-5.3-Flash | internlm/Atria-Dawn-Preview |
| Category | Open-Weight Models | Open-Weight Models |
| Official | — | Shanghai AI Laboratory ↗ |
Highlighted rows are where these differ.
GLM-5.3-Flash
- MIT licensed, which is about as permissive as an open release gets
- 320B total parameters, 18B active, with a 1M token context
- $0.15 in and $0.50 out per million tokens
Best for permissive licence at scale.
Atria Dawn Preview
- Built on Zhipu's open GLM-5.2 rather than trained from scratch
- 96.0 on DeepSearchQA and 86.5 on CyberGym, both reported by the lab
- BF16 and FP8 checkpoints, served through SGLang or vLLM
Best for agentic work on your own cluster.