OpenAI compatible API. Attested gateway. Public status.

GLM 5.2 Performance

Name: GLM 5.2 TrustedRouter performance measurements
Creator: TrustedRouter
License: https://www.apache.org/licenses/LICENSE-2.0

TrustedRouter performance signals and provider route posture for GLM 5.2.

Verify gateway

1 URLbase_url migration

100smodels and routes

0prompt logs by default

`z-ai/glm-5.2`

open weights Performance

All models

AI IQ IQ 117 #10 public AI IQ rank for glm-5.2

View AI IQ profile

Measured performance

Continuously sampled p50/p95 time-to-first-token (TTFT), time-to-first-byte (TTFB), throughput, and success rate for GLM 5.2 — 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	Throughput	Uptime	Config excluded	Samples
crusoe	2393 ms	22931 ms	2392 ms	—	100.00%	—	13
siliconflow	3265 ms	15091 ms	3264 ms	—	100.00%	—	24
parasail	3456 ms	14406 ms	3455 ms	—	100.00%	—	10
phala	4078 ms	16083 ms	4077 ms	—	83.33%	—	12
fireworks	4656 ms	19536 ms	4655 ms	—	100.00%	—	14
baseten	5469 ms	11213 ms	5468 ms	78 tok/s	93.33%	—	75
friendli	5896 ms	15532 ms	5896 ms	—	100.00%	—	6
zai	7296 ms	16369 ms	7295 ms	—	100.00%	—	28
venice	7440 ms	18122 ms	7439 ms	—	96.00%	—	25
together	7476 ms	21452 ms	7475 ms	—	100.00%	2 `probe_config_error`	56
gmi	7959 ms	22776 ms	7959 ms	—	93.55%	—	62
tinfoil	9228 ms	24553 ms	9228 ms	—	93.48%	—	46
deepinfra	10005 ms	17728 ms	10005 ms	—	92.59%	—	27
wafer	10133 ms	24058 ms	10132 ms	—	90.00%	—	10

Full provider & model leaderboard.

Provider diversity

30 routes.

More routes give the auto router more room to fail over around provider 429 and 5xx responses.

Streaming

Gateway overhead is measured separately.

Public status separates TLS/health overhead from full model latency so slow LLMs do not inflate the router metric.

Status

Metadata rollups.

Status samples store latency, outcome, provider, model, route, cost, and region metadata only.