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 |
|---|---|---|---|---|---|---|
| wafer | 1002 ms | 1219 ms | — | 100.00% | — | 13 |
| telnyx | 2000 ms | 16660 ms | — | 94.74% | — | 19 |
| friendli | 3847 ms | 5250 ms | — | 100.00% | 2 provider_error |
18 |
| privatemode | 826 ms | 2200 ms | — | 66.67% | 6 provider_error |
3 |
| makora | 1150 ms | 2638 ms | — | 100.00% | — | 3 |
| engy | 1268 ms | 1268 ms | — | 100.00% | — | 1 |
| greenference | 1281 ms | 1281 ms | — | 100.00% | 1 rate_limit_exceeded |
1 |
| inceptron | 1332 ms | 3378 ms | — | 100.00% | — | 4 |
| boundless | 1389 ms | 1389 ms | — | 100.00% | — | 1 |
| sail-research | 1397 ms | 1397 ms | — | 100.00% | — | 1 |
| wandb | 1452 ms | 3130 ms | — | 100.00% | — | 2 |
| featherless | 1736 ms | 4191 ms | — | 100.00% | — | 2 |
| cloudflare-workers-ai | 1905 ms | 1905 ms | — | 100.00% | — | 1 |
| siliconflow | 1963 ms | 1963 ms | — | 100.00% | — | 1 |
| tencent | 2414 ms | 2414 ms | — | 100.00% | — | 1 |
| zro | 2465 ms | 3607 ms | — | 100.00% | — | 5 |
| streamlake | 2521 ms | 4027 ms | — | 100.00% | — | 3 |
| baseten | 2560 ms | 2921 ms | — | 100.00% | — | 3 |
| byteplus | 2817 ms | 3578 ms | — | 100.00% | — | 4 |
| arcee | 3201 ms | 4681 ms | — | 100.00% | — | 4 |
| pearl | 3302 ms | 3444 ms | — | 100.00% | — | 3 |
| zai | 3340 ms | 4577 ms | — | 100.00% | — | 3 |
| lyceum | 3814 ms | 3814 ms | — | 100.00% | — | 1 |
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.