GLM 5.1: providers compared

Every endpoint serving this model, side by side. ◆ marks the best value in each column — per model, never fleet-wide.

Current state

	State
AtlasCloudatlas-cloud/fp8fp8	Healthy	100%◆ (best)	99.68%	99.72%	99.72%	1.49s	5.49s	54 tps	203K	$1.26 / $3.96
Crusoecrusoe/fp8fp8	Healthy	100%◆ (best)	99.74%	99.45%	99.45%	656ms	12.9s	76 tps	203K	$1.20 / $4.40
Friendlifriendli	Healthy	100%◆ (best)	99.98%	99.96%◆ (best)	99.96%◆ (best)	248ms◆ (best)	3.99s	63 tps	203K	$1.40 / $4.40
GMICloudgmicloud/fp8fp8	Healthy	100%◆ (best)	99.62%	99.59%	99.59%	1.81s	15.3s	62 tps	203K	$0.910 / $2.86
Nebiusnebius/fp8fp8	Healthy	100%◆ (best)	98.20%	97.56%	97.56%	783ms	12.0s	28 tps	203K	$1.40 / $4.40
Parasailparasail/fp8fp8	Healthy	100%◆ (best)	98.92%	99.43%	99.43%	707ms	11.0s	96 tps◆ (best)	203K	$1.40 / $4.40
SiliconFlowsiliconflow/fp8fp8	Healthy	100%◆ (best)	99.94%	99.89%	99.89%	1.70s	6.96s	48 tps	205K	$1.19 / $3.74
StreamLakestreamlake/fp8fp8	Healthy	100%◆ (best)	99.51%	98.85%	98.85%	1.92s	6.82s	29 tps	200K	$0.966 / $3.04
Z.AIz-ai/fp8fp8	Healthy	100%◆ (best)	99.92%	99.11%	99.11%	4.06s	6.69s	23 tps	203K	$1.40 / $4.40
Baidubaidu/fp8fp8	Healthy	99.65%	99.97%	99.90%	99.90%	1.55s	7.48s	52 tps	203K	$0.896 / $2.82◆ (best)
Chuteschutes/fp8fp8	Down	75.44%	76.15%	83.07%	83.07%	3.34s	74.5s	26 tps	203K	$0.980 / $3.08
Alibabaalibaba/fp8fp8	No data	—	99.90%	99.78%	99.78%	1.78s	3.91s◆ (best)	56 tps	203K	$1.33 / $4.18
DeepInfradeepinfra/fp4fp4	No data	—	99.70%	99.41%	99.41%	5.00s	7.38s	8.0 tps	203K	$1.05 / $3.50
DigitalOceandigitalocean	No data	—	97.37%	94.17%	94.17%	1.82s	17.2s	11 tps	164K	$1.30 / $4.30
Novitanovita/fp8fp8	No data	—	99.87%	99.36%	99.36%	1.76s	6.34s	59 tps	205K	$1.38 / $4.40
Phalaphala	No data	—	76.83%	83.99%	83.99%	2.35s	27.0s	28 tps	203K	$1.21 / $4.20
Venicevenice/fp8fp8	No data	—	97.86%	96.55%	96.55%	6.07s	8.21s	10 tps	200K	$1.54 / $4.84

AtlasCloud100%◆ (best)
Healthy24h 99.68%
7d
99.72%
30d
99.72%
p50
1.49s
p99
5.49s
Throughput
54 tps
Context
203K
$/Mtok
$1.26 / $3.96
atlas-cloud/fp8fp8
Crusoe100%◆ (best)
Healthy24h 99.74%
7d
99.45%
30d
99.45%
p50
656ms
p99
12.9s
Throughput
76 tps
Context
203K
$/Mtok
$1.20 / $4.40
crusoe/fp8fp8
Friendli100%◆ (best)
Healthy24h 99.98%
7d
99.96%◆ (best)
30d
99.96%◆ (best)
p50
248ms◆ (best)
p99
3.99s
Throughput
63 tps
Context
203K
$/Mtok
$1.40 / $4.40
friendli
GMICloud100%◆ (best)
Healthy24h 99.62%
7d
99.59%
30d
99.59%
p50
1.81s
p99
15.3s
Throughput
62 tps
Context
203K
$/Mtok
$0.910 / $2.86
gmicloud/fp8fp8
Nebius100%◆ (best)
Healthy24h 98.20%
7d
97.56%
30d
97.56%
p50
783ms
p99
12.0s
Throughput
28 tps
Context
203K
$/Mtok
$1.40 / $4.40
nebius/fp8fp8
Parasail100%◆ (best)
Healthy24h 98.92%
7d
99.43%
30d
99.43%
p50
707ms
p99
11.0s
Throughput
96 tps◆ (best)
Context
203K
$/Mtok
$1.40 / $4.40
parasail/fp8fp8
SiliconFlow100%◆ (best)
Healthy24h 99.94%
7d
99.89%
30d
99.89%
p50
1.70s
p99
6.96s
Throughput
48 tps
Context
205K
$/Mtok
$1.19 / $3.74
siliconflow/fp8fp8
StreamLake100%◆ (best)
Healthy24h 99.51%
7d
98.85%
30d
98.85%
p50
1.92s
p99
6.82s
Throughput
29 tps
Context
200K
$/Mtok
$0.966 / $3.04
streamlake/fp8fp8
Z.AI100%◆ (best)
Healthy24h 99.92%
7d
99.11%
30d
99.11%
p50
4.06s
p99
6.69s
Throughput
23 tps
Context
203K
$/Mtok
$1.40 / $4.40
z-ai/fp8fp8
Baidu99.65%
Healthy24h 99.97%
7d
99.90%
30d
99.90%
p50
1.55s
p99
7.48s
Throughput
52 tps
Context
203K
$/Mtok
$0.896 / $2.82◆ (best)
baidu/fp8fp8
Chutes75.44%
Down24h 76.15%
7d
83.07%
30d
83.07%
p50
3.34s
p99
74.5s
Throughput
26 tps
Context
203K
$/Mtok
$0.980 / $3.08
chutes/fp8fp8
Alibaba—
No data24h 99.90%
7d
99.78%
30d
99.78%
p50
1.78s
p99
3.91s◆ (best)
Throughput
56 tps
Context
203K
$/Mtok
$1.33 / $4.18
alibaba/fp8fp8
DeepInfra—
No data24h 99.70%
7d
99.41%
30d
99.41%
p50
5.00s
p99
7.38s
Throughput
8.0 tps
Context
203K
$/Mtok
$1.05 / $3.50
deepinfra/fp4fp4
DigitalOcean—
No data24h 97.37%
7d
94.17%
30d
94.17%
p50
1.82s
p99
17.2s
Throughput
11 tps
Context
164K
$/Mtok
$1.30 / $4.30
digitalocean
Novita—
No data24h 99.87%
7d
99.36%
30d
99.36%
p50
1.76s
p99
6.34s
Throughput
59 tps
Context
205K
$/Mtok
$1.38 / $4.40
novita/fp8fp8
Phala—
No data24h 76.83%
7d
83.99%
30d
83.99%
p50
2.35s
p99
27.0s
Throughput
28 tps
Context
203K
$/Mtok
$1.21 / $4.20
phala
Venice—
No data24h 97.86%
7d
96.55%
30d
96.55%
p50
6.07s
p99
8.21s
Throughput
10 tps
Context
200K
$/Mtok
$1.54 / $4.84
venice/fp8fp8

Measured 4m ago. Latency and throughput are OpenRouter's rolling 30-minute windows. ◆ marks the best value in each column among these endpoints.

Price vs latency

One dot per endpoint at the latest round · further left = faster, lower = cheaper · hover a dot for its tag

History

Availability

Axis starts at 90% so small differences stay visible; it expands to 0 when an endpoint falls below.

Time to first token

Logarithmic axis — the fleet spans two orders of magnitude. P50 is OpenRouter's rolling 30-minute window, averaged over the hour.

Throughput

Median tokens per second.