Best cell per row highlighted. Null means undisclosed — never counted as zero.
| Provider | Z.ai | Z.ai | Z.ai |
| Family / tier | GLM-5 · Flagship | GLM-5 · Flash | GLM-5 · Flagship |
| Status | GA | GA | GA |
| Released | 2026-06-16 | 2026-08-26 | 2026-08-14 |
| Knowledge cutoff | — | — | — |
| Context window | 1M | 1M | 1M |
| Max output tokens | 131K | 131K | 131K |
| Input modalities | text | text, image, video | text |
| Reasoning mode | optional | always | always |
| Open weights | Yes | Yes | Yes |
| Input | $1.4 | $0.15 | $1.4 |
| Output | $4.4 | $0.5 | $4.4 |
| Cached input | $0.26 | $0.03 | $0.26 |
| Batch input | — | — | — |
| Free tier | No | No | No |
| Output speed (tok/s) | 166.6 | — | — |
| Time to first token (s) | 1.39 | — | — |
| Latency tier | fast | medium | medium |
| Trains on inputs | — | No | No |
| License | MIT | MIT | custom-GLM-5.3-License |
| Clouds | — | — | Mistral AI, Alibaba Cloud Model Studio |
| Compliance | — | — | — |
| Unified score | 8.8 | 8.6 | 8.7 |
| Decision Maker | 8.5 | 8.5 | 8.5 |
| Domain Strategist | 9 | 9 | 9 |
| Finance Lead | 9.5 | 9.5 | 9.5 |
| Domain Practitioner | 9 | 9 | 9 |
| Power User | 8.5 | 8.5 | 8.5 |
| Skeptic | 8 | 7 | 7.5 |
| Value score | 9.6 | 9.8 | 9.6 |