Alibaba vs 
Zhipu AI
Verdict
- Their flagships are level: GLM 5.3 63.5 vs Qwen3.8 Max 63.3.
- Zhipu AI has the better model in 4 of 4 price bands.
- Alibaba's fastest ranked model (Qwen3 Coder 30B A3B Instruct, 244 tok/s) beats Zhipu AI's (GLM 5.2, 94 tok/s).
- Alibaba publishes more open-weights models (48 vs 15).
- Alibaba ships more often: one model every 7 days against every 28.
Best model per price band
Each provider’s highest-index model below the price ceiling. Purple marks the winner of the band.
| Price band | Alibaba | Zhipu AI |
|---|---|---|
| Under $0.50/M | Qwen3.8-Flash-Next 60.3$0.230/M | GLM 5.3 Flash 62.7$0.238/M |
| Under $2/M | Qwen3.8-Flash-Next 60.3$0.230/M | GLM 5.3 Flash 62.7$0.238/M |
| Under $10/M | Qwen3.8 Max max 63.3$3.00/M | GLM 5.3 max 63.5$2.15/M |
| Any price | Qwen3.8 Max max 63.3$3.00/M | GLM 5.3 max 63.5$2.15/M |
Line-up
| Metric | Alibaba | Zhipu AI |
|---|---|---|
| Flagship index | 63.3 | 63.5 |
| Ranked models | 54 | 14 |
| Open-weights models | 48 | 15 |
| Cheapest ranked, $/M | $0.060 | $0.153 |
| Fastest ranked, tok/s | 244 | 94 |
| Release cadence, days | 7 | 28 |
| Last release | 17 Sept 2026 | 18 Sept 2026 |
| Next expected | 21 Sept 2026 – 26 Sept 2026 | 4 Oct 2026 – 27 Oct 2026 |