DeepSeek V4 Flash 0423: 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
Cloudflarecloudflare	Healthy	100%◆ (best)	99.14%	99.71%	99.71%	1.59s	15.5s	65 tps	384K	$0.440 / $1.32
GMICloudgmicloud/fp8fp8	Healthy	100%◆ (best)	99.92%	99.97%◆ (best)	99.69%	1.68s	5.52s◆ (best)	53 tps	1.0M	$0.091 / $0.182
SiliconFlowsiliconflow/fp8fp8	Healthy	100%◆ (best)	99.41%	99.73%	98.28%	1.10s	8.01s	38 tps	1.0M	$0.130 / $0.280
Novitanovita/fp8fp8	Healthy	99.96%	99.96%	99.95%	99.93%◆ (best)	1.51s	5.60s	55 tps	1.0M	$0.140 / $0.280
Relacerelace/fp4fp4	Healthy	99.94%	99.92%	99.64%	99.73%	569ms◆ (best)	7.49s	104 tps◆ (best)	1.0M	$0.030 / $1.28
DigitalOceandigitalocean	Healthy	99.89%	99.80%	99.82%	99.83%	1.02s	19.2s	14 tps	1.0M	$0.098 / $0.196
StreamLakestreamlake/fp8fp8	Healthy	99.87%	98.99%	99.04%	97.67%	1.49s	21.0s	51 tps	1.0M	$0.042 / $0.084
AtlasCloudatlas-cloud/fp4fp4	Healthy	99.84%	98.77%	98.41%	99.26%	1.63s	10.2s	54 tps	1.0M	$0.140 / $0.280
Parasailparasail/fp8fp8	Healthy	99.76%	99.74%	99.80%	99.72%	791ms	7.23s	65 tps	1.0M	$0.140 / $0.280
OpenInferenceopen-inference/fp4fp4	Healthy	99.06%	99.58%	98.62%	98.62%	841ms	8.47s	34 tps	1.0M	$0.039 / $1.66
Alibabaalibaba/fp8fp8	Healthy	98.89%	98.98%	98.78%	99.10%	855ms	5.84s	94 tps	1.0M	$0.134 / $0.268
Baidubaidu/fp8fp8	Healthy	98.12%	99.56%	89.53%	96.37%	1.06s	12.7s	67 tps	1.0M	$0.140 / $0.280
Azureazure/us	Degraded	97.79%	96.90%	97.61%	95.91%	1.24s	24.1s	54 tps	1.0M	$0.210 / $0.560
Venicevenice	Degraded	94.96%	98.31%	98.91%	98.40%	1.19s	15.6s	51 tps	1.0M	$0.097 / $0.193
Mancer 2mancer/fp8fp8	Down	81.72%	96.63%	97.28%	97.95%	1.13s	16.2s	19 tps	1.0M	$0.190 / $0.500
CoreWeavecoreweave/fp8fp8	No data	—	—	—	—	—	—	—	1.0M	$0.140 / $0.280
Crusoecrusoe/fp8fp8	No data	—	—	—	—	—	—	—	1.0M	$0.140 / $0.280
DeepInfradeepinfra/fp8fp8	No data	—	99.82%	99.55%	99.49%	9.31s	21.8s	10 tps	1.0M	$0.090 / $0.180
DeepSeekdeepseek	No data	—	—	—	—	—	—	—	1.0M	$0.220 / $0.660
NextBitnextbit/fp8fp8	No data	—	—	—	99.35%	—	—	—	1.0M	$0.150 / $0.300
OpenInferenceopen-inference/fp8fp8	No data	—	—	95.36%	97.88%	—	—	—	1.0M	$0.002 / $1.31◆ (best)
Phalaphala	No data	—	—	—	99.71%	—	—	—	1.0M	$0.200 / $0.400
Rekareka	No data	—	—	98.07%	98.07%	—	—	—	1.0M	$0.130 / $0.260
Sail Researchsail-research/fp4fp4	No data	—	—	—	—	—	—	—	1.0M	$0.090 / $0.180
Waferwafer	No data	—	—	—	99.43%	—	—	—	1.0M	$0.100 / $0.250

Cloudflare100%◆ (best)
Healthy24h 99.14%
7d
99.71%
30d
99.71%
p50
1.59s
p99
15.5s
Throughput
65 tps
Context
384K
$/Mtok
$0.440 / $1.32
cloudflare
GMICloud100%◆ (best)
Healthy24h 99.92%
7d
99.97%◆ (best)
30d
99.69%
p50
1.68s
p99
5.52s◆ (best)
Throughput
53 tps
Context
1.0M
$/Mtok
$0.091 / $0.182
gmicloud/fp8fp8
SiliconFlow100%◆ (best)
Healthy24h 99.41%
7d
99.73%
30d
98.28%
p50
1.10s
p99
8.01s
Throughput
38 tps
Context
1.0M
$/Mtok
$0.130 / $0.280
siliconflow/fp8fp8
Novita99.96%
Healthy24h 99.96%
7d
99.95%
30d
99.93%◆ (best)
p50
1.51s
p99
5.60s
Throughput
55 tps
Context
1.0M
$/Mtok
$0.140 / $0.280
novita/fp8fp8
Relace99.94%
Healthy24h 99.92%
7d
99.64%
30d
99.73%
p50
569ms◆ (best)
p99
7.49s
Throughput
104 tps◆ (best)
Context
1.0M
$/Mtok
$0.030 / $1.28
relace/fp4fp4
DigitalOcean99.89%
Healthy24h 99.80%
7d
99.82%
30d
99.83%
p50
1.02s
p99
19.2s
Throughput
14 tps
Context
1.0M
$/Mtok
$0.098 / $0.196
digitalocean
StreamLake99.87%
Healthy24h 98.99%
7d
99.04%
30d
97.67%
p50
1.49s
p99
21.0s
Throughput
51 tps
Context
1.0M
$/Mtok
$0.042 / $0.084
streamlake/fp8fp8
AtlasCloud99.84%
Healthy24h 98.77%
7d
98.41%
30d
99.26%
p50
1.63s
p99
10.2s
Throughput
54 tps
Context
1.0M
$/Mtok
$0.140 / $0.280
atlas-cloud/fp4fp4
Parasail99.76%
Healthy24h 99.74%
7d
99.80%
30d
99.72%
p50
791ms
p99
7.23s
Throughput
65 tps
Context
1.0M
$/Mtok
$0.140 / $0.280
parasail/fp8fp8
OpenInference99.06%
Healthy24h 99.58%
7d
98.62%
30d
98.62%
p50
841ms
p99
8.47s
Throughput
34 tps
Context
1.0M
$/Mtok
$0.039 / $1.66
open-inference/fp4fp4
Alibaba98.89%
Healthy24h 98.98%
7d
98.78%
30d
99.10%
p50
855ms
p99
5.84s
Throughput
94 tps
Context
1.0M
$/Mtok
$0.134 / $0.268
alibaba/fp8fp8
Baidu98.12%
Healthy24h 99.56%
7d
89.53%
30d
96.37%
p50
1.06s
p99
12.7s
Throughput
67 tps
Context
1.0M
$/Mtok
$0.140 / $0.280
baidu/fp8fp8
Azure97.79%
Degraded24h 96.90%
7d
97.61%
30d
95.91%
p50
1.24s
p99
24.1s
Throughput
54 tps
Context
1.0M
$/Mtok
$0.210 / $0.560
azure/us
Venice94.96%
Degraded24h 98.31%
7d
98.91%
30d
98.40%
p50
1.19s
p99
15.6s
Throughput
51 tps
Context
1.0M
$/Mtok
$0.097 / $0.193
venice
Mancer 281.72%
Down24h 96.63%
7d
97.28%
30d
97.95%
p50
1.13s
p99
16.2s
Throughput
19 tps
Context
1.0M
$/Mtok
$0.190 / $0.500
mancer/fp8fp8
CoreWeave—
No data24h —
7d
—
30d
—
p50
—
p99
—
Throughput
—
Context
1.0M
$/Mtok
$0.140 / $0.280
coreweave/fp8fp8
Crusoe—
No data24h —
7d
—
30d
—
p50
—
p99
—
Throughput
—
Context
1.0M
$/Mtok
$0.140 / $0.280
crusoe/fp8fp8
DeepInfra—
No data24h 99.82%
7d
99.55%
30d
99.49%
p50
9.31s
p99
21.8s
Throughput
10 tps
Context
1.0M
$/Mtok
$0.090 / $0.180
deepinfra/fp8fp8
DeepSeek—
No data24h —
7d
—
30d
—
p50
—
p99
—
Throughput
—
Context
1.0M
$/Mtok
$0.220 / $0.660
deepseek
NextBit—
No data24h —
7d
—
30d
99.35%
p50
—
p99
—
Throughput
—
Context
1.0M
$/Mtok
$0.150 / $0.300
nextbit/fp8fp8
OpenInference—
No data24h —
7d
95.36%
30d
97.88%
p50
—
p99
—
Throughput
—
Context
1.0M
$/Mtok
$0.002 / $1.31◆ (best)
open-inference/fp8fp8
Phala—
No data24h —
7d
—
30d
99.71%
p50
—
p99
—
Throughput
—
Context
1.0M
$/Mtok
$0.200 / $0.400
phala
Reka—
No data24h —
7d
98.07%
30d
98.07%
p50
—
p99
—
Throughput
—
Context
1.0M
$/Mtok
$0.130 / $0.260
reka
Sail Research—
No data24h —
7d
—
30d
—
p50
—
p99
—
Throughput
—
Context
1.0M
$/Mtok
$0.090 / $0.180
sail-research/fp4fp4
Wafer—
No data24h —
7d
—
30d
99.43%
p50
—
p99
—
Throughput
—
Context
1.0M
$/Mtok
$0.100 / $0.250
wafer

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.

Throughput

Median tokens per second.