Z.ai: GLM 5.3 Flash Performance
Try GLM 5.3 Flash on us. Starter credit covers your first requests, then one key reaches all 32 provider routes with automatic failover.
z-ai/glm-5.3-flash
Measured performance
Continuously sampled p50/p95 time-to-first-token (TTFT), effective throughput, and success rate for Z.ai: GLM 5.3 Flash. 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 | Effective throughput | Uptime | Config excluded | Availability samples |
|---|---|---|---|---|---|---|
| together | 930 ms | 1054 ms | — | 100.00% | — | 3 |
| wafer | 1162 ms | 1162 ms | — | 100.00% | — | 1 |
| wandb | 1173 ms | 1173 ms | — | 100.00% | — | 1 |
| makora | 1273 ms | 2035 ms | — | 100.00% | — | 4 |
| parasail | 1312 ms | 1312 ms | — | 100.00% | — | 1 |
| inceptron | 1329 ms | 1329 ms | — | 100.00% | 4 insufficient_quota |
1 |
| baseten | 1347 ms | 1347 ms | — | 100.00% | — | 1 |
| zro | 1412 ms | 1582 ms | — | 100.00% | — | 2 |
| greenference | 1430 ms | 1723 ms | — | 100.00% | — | 3 |
| sail-research | 1500 ms | 1500 ms | — | 100.00% | — | 1 |
| atlas-cloud | 1503 ms | 1503 ms | — | 100.00% | — | 1 |
| zai | 1611 ms | 2304 ms | — | 100.00% | — | 2 |
| arcee | 1619 ms | 2093 ms | — | 100.00% | — | 2 |
| engy | 1688 ms | 1688 ms | — | 100.00% | — | 1 |
| siliconflow | 1790 ms | 2448 ms | — | 100.00% | — | 2 |
| deepinfra | 2375 ms | 2375 ms | — | 100.00% | — | 1 |
| novita | 2572 ms | 2572 ms | — | 100.00% | — | 1 |
| byteplus | 2786 ms | 2883 ms | — | 100.00% | — | 2 |
| tencent | 2860 ms | 2860 ms | — | 100.00% | — | 1 |
| pearl | 2944 ms | 3435 ms | — | 100.00% | — | 3 |
| streamlake | 4254 ms | 4254 ms | — | 100.00% | — | 1 |
| privatemode | — | — | — | 0.00% | 2 provider_auth_config |
5 |
Full provider & model leaderboard.
32 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.