Z.ai: GLM 5.2 Performance
TrustedRouter performance signals and provider route posture for Z.ai: GLM 5.2.
z-ai/glm-5.2
Measured performance
Continuously sampled p50/p95 time-to-first-token (TTFT), time-to-first-byte (TTFB), effective throughput, and success rate for Z.ai: GLM 5.2. Effective throughput uses provider-reported output tokens over complete request time. Unsupported route and probe-configuration rows are separated from provider downtime, and no prompt or output content is stored.
| Provider | p50 TTFT | p95 TTFT | p50 TTFB | Effective throughput | Uptime | Config excluded | Availability samples |
|---|---|---|---|---|---|---|---|
| friendli | 1350 ms | 3348 ms | 1350 ms | 94 tok/s n=2 | 100.00% | — | 13 |
| telnyx | 1454 ms | 3647 ms | 1454 ms | 100 tok/s n=2 | 100.00% | — | 6 |
| siliconflow | 1729 ms | 3646 ms | 1729 ms | 30 tok/s n=2 | 100.00% | — | 3 |
| atlas-cloud | 1860 ms | 3024 ms | 1860 ms | 46 tok/s n=1 | 100.00% | — | 2 |
| morph | 1930 ms | 14247 ms | 1930 ms | 53 tok/s n=2 | 100.00% | — | 8 |
| makora | 2148 ms | 4159 ms | 2147 ms | — | 100.00% | — | 7 |
| zai | 2222 ms | 3949 ms | 2222 ms | 67 tok/s n=2 | 100.00% | — | 3 |
| together | 2263 ms | 18324 ms | 2263 ms | — | 100.00% | — | 7 |
| phala | 2361 ms | 5124 ms | 2361 ms | 49 tok/s n=2 | 100.00% | — | 27 |
| wafer | 2426 ms | 5346 ms | 2425 ms | 190 tok/s n=2 | 100.00% | — | 12 |
| crusoe | 2584 ms | 4984 ms | 2584 ms | 154 tok/s n=2 | 100.00% | — | 4 |
| parasail | 2730 ms | 12604 ms | 984 ms | 62 tok/s n=2 | 100.00% | — | 13 |
| fireworks | 2869 ms | 4542 ms | 2868 ms | 43 tok/s n=2 | 100.00% | — | 3 |
| baseten | 2912 ms | 3474 ms | 2912 ms | 45 tok/s n=2 | 100.00% | — | 5 |
| venice | 3492 ms | 17844 ms | 3492 ms | 33 tok/s n=2 | 100.00% | — | 5 |
| chutes | 3914 ms | 9240 ms | 3914 ms | 31 tok/s n=2 | 100.00% | — | 4 |
| gmi | 3941 ms | 5872 ms | 3941 ms | 50 tok/s n=2 | 100.00% | — | 18 |
| inceptron | 5987 ms | 29961 ms | 5987 ms | — | 100.00% | — | 16 |
| tinfoil | 3151 ms | 46057 ms | 3151 ms | — | 85.71% | — | 14 |
| digitalocean | 5351 ms | 5351 ms | 5351 ms | 28 tok/s n=2 | 50.00% | — | 2 |
| alibaba | — | — | — | 37 tok/s n=2 | — | — | 0 |
| deepinfra | — | — | — | 49 tok/s n=2 | — | — | 0 |
| novita | — | — | — | 25 tok/s n=2 | — | — | 0 |
Full provider & model leaderboard.
43 routes.
More routes give the auto router more room to fail over around provider 429 and 5xx responses.
Gateway overhead is measured separately.
Public status separates TLS/health overhead from full model latency so slow LLMs do not inflate the router metric.
Metadata rollups.
Status samples store latency, outcome, provider, model, route, cost, and region metadata only.
View public status or inspect provider routes.