zai-org/GLM-5.3-Flash
Parameters321B
Context1M
Input → outputimage + text → text
Licensemit
PrecisionF8_E4M3 · fp8
ArchitectureGlm5NextForConditionalGeneration
ReleasedAug 25, 2026
Downloads6.5M
Other sizes and formats in the family
| Model | Params | Precision | Downloads |
|---|---|---|---|
| zai-org/GLM-5.3-Flash-BF16 | 321B | BF16 | 70.9K |
Built on this model
| Model | Params | Precision | Downloads |
|---|---|---|---|
| nvidia/GLM-5.3-Flash-NVFP4nvfp4 | 169B | U8 | 92.9K |
This release in the news
No coverage of this exact release yet.
Recent news about GLM
- Oct 8, 2026 · Research paperSafe Actions Alone Do Not Ensure Safe Agents: Identifying Unfulfilled Obligations with Guard Models
- Oct 8, 2026 · Research paperREMORY: Learning Residual Memory for Context Compaction
- Oct 8, 2026 · Research paperQUILT: Rethinking Sparse-Attention Prefill through Shared Query Execution
- Oct 7, 2026 · Research paperHow Narrative Wrapping Affects LLM Refusal: A Cross-Language Benchmark and Defense
- Oct 6, 2026 · Research paperSocio-Foundation: A Model for Generalizable Individual Behavior Simulation via Hierarchical Capability Distillation
- Oct 6, 2026 · Research paperMINDSET: Energy-based Schema Evolution for Long Conversational Agent Memory
- Oct 6, 2026 · Opinion / analysisThe Cyber Risk Discourse is Broken
- Oct 6, 2026 · Research paperImage Bitstream Fine-grained Understanding for Privacy-Friendly AIoT