SpaceXAI vs 
Zhipu AI
Verdict
- Their flagships are level: Grok 4.6 66.0 vs GLM 5.3 65.8.
- SpaceXAI's fastest ranked model (Grok 4.20 Multi-Agent, 304 tok/s) beats Zhipu AI's (GLM 4.7, 106 tok/s).
- Zhipu AI publishes more open-weights models (18 vs 1).
- Zhipu AI ships more often: one model every 28 days against every 42.
Best model per price band
Each provider’s highest-index model below the price ceiling. Purple marks the winner of the band.
| Price band | SpaceXAI | Zhipu AI |
|---|---|---|
| Under $0.50/M | – | GLM 5.3 Flash 63.9$0.238/M |
| Under $2/M | Grok 4.20 thinking 59.9$1.56/M | GLM 5.3 Flash 63.9$0.238/M |
| Under $10/M | Grok 4.6 medium 66.0$3.00/M | GLM 5.3 max 65.8$2.15/M |
| Any price | Grok 4.6 medium 66.0$3.00/M | GLM 5.3 max 65.8$2.15/M |
Line-up
| Metric | SpaceXAI | Zhipu AI |
|---|---|---|
| Flagship index | 66.0 | 65.8 |
| Ranked models | 16 | 14 |
| Open-weights models | 1 | 18 |
| Cheapest ranked, $/M | $1.25 | $0.153 |
| Fastest ranked, tok/s | 304 | 106 |
| Release cadence, days | 42 | 28 |
| Last release | 21 Sept 2026 | 18 Sept 2026 |
| Next expected | 16 Oct 2026 – 18 Nov 2026 | 4 Oct 2026 – 27 Oct 2026 |