Z.ai: GLM 4.5 Air
z-ai/glm-4.5-air
Cheapest provider
$0.13 / 1M
Io Net
Fastest provider (p95)
85 tok/s
Io Net
Intelligence (estimated)
68.5
Family + generation + popularity + price tier
Per-provider performance
Latency / throughput / uptime measured across providers over the last 30 minutes of live traffic. Atlas’s router weighs these per call (with the eval-gate signal) when picking a variant for Standard and Batch tiers.
| Provider | Quant | Input $/1M | Output $/1M | Latency p50 / p95 | Throughput p50 / p95 | Uptime 30m | Success |
|---|---|---|---|---|---|---|---|
| Io Net | q8· fp8 | $0.1250 | $0.8400 | 775ms / 3777ms | 35 / 85 tok/s | 99.91% | 99.7% |
| Novita | full· bf16 | $0.1300 | $0.8500 | 814ms / 2482ms | 38 / 54 tok/s | 100.00% | 99.4% |
| SiliconFlow | q8· fp8 | $0.1400 | $0.8600 | 2224ms / 6768ms | 12 / 54.1 tok/s | 100.00% | 100.0% |
| Z.AI | q8· fp8 | $0.2000 | $1.1000 | 2473ms / 6488ms | 35 / 54 tok/s | 99.97% | 100.0% |
“—” means live telemetry hasn’t accumulated enough recent traffic for that endpoint. “undisclosed” means the provider serves the model but doesn’t expose the quantization label (typically running fp8 / int8 internally).
Intelligence estimate
No public benchmark numbers indexed for this model yet, so the leaderboard score is derived from the catalogue data we sync: model family standing, generation, and live usage rank — applied identically to open- and closed-weight models. The heuristic ceiling sits below confirmed-benchmark frontier models so curated rankings stay clearly on top.
- ·glm air family base
- ·v4.5 generation +3
Want a curated score? File a benchmark report and we’ll add it to the next sync.
How Atlas routes Z.ai: GLM 4.5 Air
- Realtime tier — direct passthrough at the upstream’s native precision. Best for hard latency / quality guarantees.
- Standard tier — Atlas picks the cheapest provider variant whose quantization has stayed green on your operation’s eval gates. For Z.ai: GLM 4.5 Air that’s currently Io Net at $0.13/1M.
- Batch tier — async, biggest discount. Roadmapped to use provider batch APIs (OpenAI / Anthropic) where available and queued spot capacity for open-weight workloads.