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 |
|---|---|---|---|---|---|---|---|
| siliconflow | 1308 ms | 1729 ms | 1308 ms | 29 tok/s n=1 | 100.00% | — | 2 |
| friendli | 1545 ms | 3348 ms | 1544 ms | 75 tok/s n=1 | 100.00% | — | 11 |
| fireworks | 2007 ms | 2850 ms | 2007 ms | 38 tok/s n=1 | 100.00% | — | 2 |
| makora | 2148 ms | 24352 ms | 2147 ms | — | 100.00% | — | 7 |
| zai | 2195 ms | 3949 ms | 2195 ms | 63 tok/s n=1 | 100.00% | — | 5 |
| morph | 2213 ms | 14247 ms | 2213 ms | 50 tok/s n=1 | 100.00% | — | 8 |
| together | 2263 ms | 18324 ms | 2263 ms | — | 100.00% | — | 8 |
| wafer | 2365 ms | 5346 ms | 2364 ms | 168 tok/s n=1 | 100.00% | — | 6 |
| phala | 2641 ms | 5124 ms | 2640 ms | 59 tok/s n=1 | 100.00% | — | 27 |
| parasail | 2766 ms | 12604 ms | — | 60 tok/s n=1 | 100.00% | — | 11 |
| baseten | 2912 ms | 3474 ms | 2912 ms | 56 tok/s n=1 | 100.00% | — | 6 |
| telnyx | 2950 ms | 3647 ms | 2950 ms | 138 tok/s n=1 | 100.00% | — | 5 |
| atlas-cloud | 3024 ms | 3024 ms | 3024 ms | — | 100.00% | — | 1 |
| venice | 3491 ms | 4987 ms | 3491 ms | 34 tok/s n=1 | 100.00% | — | 3 |
| gmi | 4038 ms | 19314 ms | 4037 ms | 56 tok/s n=1 | 100.00% | — | 16 |
| chutes | 4049 ms | 9240 ms | 4049 ms | 21 tok/s n=1 | 100.00% | — | 5 |
| crusoe | 4316 ms | 26272 ms | 4316 ms | 146 tok/s n=1 | 100.00% | — | 4 |
| inceptron | 6412 ms | 28834 ms | 6412 ms | — | 100.00% | 1 router_error |
16 |
| tinfoil | 3151 ms | 46057 ms | 3151 ms | — | 83.33% | — | 12 |
| digitalocean | 5351 ms | 9801 ms | 5351 ms | 41 tok/s n=1 | 66.67% | — | 3 |
| alibaba | — | — | — | 31 tok/s n=1 | — | — | 0 |
| deepinfra | — | — | — | 42 tok/s n=1 | — | — | 0 |
| novita | — | — | — | 35 tok/s n=1 | — | — | 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.