Our take
Z.AI is a specialist host offering ten variants of the GLM family of language models, with pricing between budget and mid-market levels and modest throughput speeds. It is useful when you need a specific GLM variant that is hard to find elsewhere.
Pick this when no other host carries the specific GLM variant you need. Consider it for mid-budget scenarios where GLM-4-5v's 59 TPS at $0.60 input and $1.80 output per million tokens offers acceptable speed-per-dollar. Skip it if your risk assessment requires verified headquarters location, SOC 2, EU data residency, or a clear training-on-prompts policy.
- Ten distinct tracked GLM models from GLM-4-5-Air through GLM-5-2 — unusual variant depth for a specialist host.
- GLM-4-5v at 59 TPS is three times faster than the catalogue's slowest entries at 19 TPS.
- Very thin transparency and compliance documentation: headquarters country, EU endpoint, SOC 2, training policy, and zero-retention are all unverified in our data.
- No price leadership at any tier. Even the cheapest offer, GLM-4-5-Air at $0.20 input and $1.10 output per million tokens, is not the market floor, and most models cluster at $0.60 or above for input.
- Inconsistent throughput with several slow options: four of ten models are at or below 27 TPS, and two run at 19 TPS despite mid-market pricing.
Privacy & data handling
- ✓
- Yes
- ✗
- No
- No answer on record
The colour says whether the answer favours you, not whether it is a yes — not training on your prompts earns a green cross, no zero-retention option earns a red one.
Prompt retention: none.