Side-by-side comparison of GLM-4.5 (Zhipu AI · China) and Jamba 1.5 Large (AI21 Labs · Israel) for self-hosted deployment of the open-weight model. GLM-4.5 is rated conditional; Jamba 1.5 Large is conditional. They part ways on licence: GLM-4.5 is "MIT", Jamba 1.5 Large is "Jamba Open".
| Field | ||
|---|---|---|
| Summary | ||
| Verdict | Conditional MoE flagship from Zhipu under MIT. Strong agentic and coding benchmarks. Same Chinese-origin alignment and geopolitical considerations as DeepSeek / Qwen. | Conditional SSM-Transformer hybrid (Mamba) with 256K context. Jamba Open Model License permits commercial use below $50M annual revenue; above that, paid licence required. Israel jurisdiction; EU adequacy decision in place. |
| Last reviewed | 2026-04-15 | 2026-04-15 |
| Open-weight | ||
| Licence | MIT | Jamba Open |
| Commercial use | Yes | Under $50M rev. |
| Training data | Undisclosed | Undisclosed |
| Origin | China | Israel |
| Performance & pricing? | ||
| Quality index | 26/100 | 11/100 |
| Speed | 44 tok/s | — |
| Blended price | $0.84/M | $3.50/M |
| Context window | — | — |
| Evidence | ||
| Sources | ||
No overlapping sources between the two entries.