DeepSeek V4 Flash 0423

deepseek/deepseek-v4-flash

Compare with DeepSeek V4 Flash 0731, DeepSeek V4 Pro 0423, DeepSeek V4 Pro 0813

CompareDeepSeek V4 Flash 0423

	State
Baidubaidu/fp8fp8	Healthy	100%	99.93%	98.66%	98.66%	935ms	4.23s	71 tps	1.0M	$0.140 / $0.280
Cloudflarecloudflare	Healthy	100%	100%	100%	100%	716ms	1.74s	66 tps	384K	$0.440 / $1.32
Phalaphala	Healthy	100%	91.75%	95.00%	95.00%	4.22s	28.7s	4.0 tps	1.0M	$0.200 / $0.400
DeepSeekdeepseek	Healthy	99.99%	99.04%	83.77%	83.77%	811ms	3.31s	82 tps	1.0M	$0.440 / $1.32
Novitanovita/fp8fp8	Healthy	99.98%	99.72%	99.75%	99.75%	1.11s	6.88s	75 tps	1.0M	$0.140 / $0.280
Alibabaalibaba/fp8fp8	Healthy	99.92%	99.79%	99.69%	99.69%	832ms	5.92s	70 tps	1.0M	$0.134 / $0.268
CoreWeavecoreweave/fp8fp8	Healthy	99.86%	99.48%	99.55%	99.55%	530ms	11.3s	35 tps	1.0M	$0.140 / $0.280
SiliconFlowsiliconflow/fp8fp8	Healthy	99.85%	99.25%	98.95%	98.95%	1.21s	9.16s	65 tps	1.0M	$0.130 / $0.280
Parasailparasail/fp8fp8	Healthy	99.79%	99.45%	99.29%	99.29%	701ms	14.2s	60 tps	1.0M	$0.140 / $0.280
StreamLakestreamlake/fp8fp8	Healthy	99.72%	99.66%	99.19%	99.19%	2.98s	13.3s	36 tps	1.0M	$0.083 / $0.165
Venicevenice	Healthy	99.68%	99.86%	99.48%	99.48%	1.65s	4.66s	36 tps	1.0M	$0.138 / $0.275
Sail Researchsail-research/fp4fp4	Healthy	99.58%	99.73%	99.70%	99.70%	755ms	6.62s	23 tps	1.0M	$0.090 / $0.180
AtlasCloudatlas-cloud/fp4fp4	Healthy	99.50%	99.88%	99.90%	99.90%	1.08s	9.54s	51 tps	1.0M	$0.140 / $0.280
DigitalOceandigitalocean	Healthy	99.46%	99.30%	98.79%	98.79%	1.74s	19.4s	6.0 tps	1.0M	$0.068 / $0.168
DeepInfradeepinfra/fp8fp8	Healthy	99.43%	99.64%	99.27%	99.27%	675ms	3.35s	38 tps	1.0M	$0.090 / $0.180
Mancer 2mancer/fp8fp8	Healthy	99.39%	98.69%	98.70%	98.70%	915ms	8.26s	37 tps	1.0M	$0.140 / $0.450
GMICloudgmicloud/fp8fp8	Healthy	99.07%	99.71%	99.38%	99.38%	2.37s	9.63s	53 tps	1.0M	$0.084 / $0.168
Azureazure/us	Degraded	—	87.23%	85.35%	85.35%	2.29s	77.0s	30 tps	1.0M	$0.210 / $0.560

Baidu100%
Healthy24h 99.93%
7d
98.66%
30d
98.66%
p50
935ms
p99
4.23s
Throughput
71 tps
Context
1.0M
$/Mtok
$0.140 / $0.280
baidu/fp8fp8
Cloudflare100%
Healthy24h 100%
7d
100%
30d
100%
p50
716ms
p99
1.74s
Throughput
66 tps
Context
384K
$/Mtok
$0.440 / $1.32
cloudflare
Phala100%
Healthy24h 91.75%
7d
95.00%
30d
95.00%
p50
4.22s
p99
28.7s
Throughput
4.0 tps
Context
1.0M
$/Mtok
$0.200 / $0.400
phala
DeepSeek99.99%
Healthy24h 99.04%
7d
83.77%
30d
83.77%
p50
811ms
p99
3.31s
Throughput
82 tps
Context
1.0M
$/Mtok
$0.440 / $1.32
deepseek
Novita99.98%
Healthy24h 99.72%
7d
99.75%
30d
99.75%
p50
1.11s
p99
6.88s
Throughput
75 tps
Context
1.0M
$/Mtok
$0.140 / $0.280
novita/fp8fp8
Alibaba99.92%
Healthy24h 99.79%
7d
99.69%
30d
99.69%
p50
832ms
p99
5.92s
Throughput
70 tps
Context
1.0M
$/Mtok
$0.134 / $0.268
alibaba/fp8fp8
CoreWeave99.86%
Healthy24h 99.48%
7d
99.55%
30d
99.55%
p50
530ms
p99
11.3s
Throughput
35 tps
Context
1.0M
$/Mtok
$0.140 / $0.280
coreweave/fp8fp8
SiliconFlow99.85%
Healthy24h 99.25%
7d
98.95%
30d
98.95%
p50
1.21s
p99
9.16s
Throughput
65 tps
Context
1.0M
$/Mtok
$0.130 / $0.280
siliconflow/fp8fp8
Parasail99.79%
Healthy24h 99.45%
7d
99.29%
30d
99.29%
p50
701ms
p99
14.2s
Throughput
60 tps
Context
1.0M
$/Mtok
$0.140 / $0.280
parasail/fp8fp8
StreamLake99.72%
Healthy24h 99.66%
7d
99.19%
30d
99.19%
p50
2.98s
p99
13.3s
Throughput
36 tps
Context
1.0M
$/Mtok
$0.083 / $0.165
streamlake/fp8fp8
Venice99.68%
Healthy24h 99.86%
7d
99.48%
30d
99.48%
p50
1.65s
p99
4.66s
Throughput
36 tps
Context
1.0M
$/Mtok
$0.138 / $0.275
venice
Sail Research99.58%
Healthy24h 99.73%
7d
99.70%
30d
99.70%
p50
755ms
p99
6.62s
Throughput
23 tps
Context
1.0M
$/Mtok
$0.090 / $0.180
sail-research/fp4fp4
AtlasCloud99.50%
Healthy24h 99.88%
7d
99.90%
30d
99.90%
p50
1.08s
p99
9.54s
Throughput
51 tps
Context
1.0M
$/Mtok
$0.140 / $0.280
atlas-cloud/fp4fp4
DigitalOcean99.46%
Healthy24h 99.30%
7d
98.79%
30d
98.79%
p50
1.74s
p99
19.4s
Throughput
6.0 tps
Context
1.0M
$/Mtok
$0.068 / $0.168
digitalocean
DeepInfra99.43%
Healthy24h 99.64%
7d
99.27%
30d
99.27%
p50
675ms
p99
3.35s
Throughput
38 tps
Context
1.0M
$/Mtok
$0.090 / $0.180
deepinfra/fp8fp8
Mancer 299.39%
Healthy24h 98.69%
7d
98.70%
30d
98.70%
p50
915ms
p99
8.26s
Throughput
37 tps
Context
1.0M
$/Mtok
$0.140 / $0.450
mancer/fp8fp8
GMICloud99.07%
Healthy24h 99.71%
7d
99.38%
30d
99.38%
p50
2.37s
p99
9.63s
Throughput
53 tps
Context
1.0M
$/Mtok
$0.084 / $0.168
gmicloud/fp8fp8
Azure—
Degraded24h 87.23%
7d
85.35%
30d
85.35%
p50
2.29s
p99
77.0s
Throughput
30 tps
Context
1.0M
$/Mtok
$0.210 / $0.560
azure/us

Measured 5m ago. Latency and throughput are OpenRouter's rolling 30-minute windows.

Price changes

DeepSeek V4 Flash 0423 DeepSeekdeepseek$0.220 / $0.660 → $0.440 / $1.322026-08-21 06:10 UTC
DeepSeek V4 Flash 0423 DeepSeekdeepseek$0.440 / $1.32 → $0.220 / $0.6602026-08-21 04:10 UTC
DeepSeek V4 Flash 0423 DeepSeekdeepseek$0.220 / $0.660 → $0.440 / $1.322026-08-21 01:10 UTC
DeepSeek V4 Flash 0423 StreamLakestreamlake/fp8$0.076 / $0.151 → $0.083 / $0.1652026-08-20 23:45 UTC
DeepSeek V4 Flash 0423 Baidubaidu/fp8$0.077 / $0.154 → $0.140 / $0.2802026-08-20 23:35 UTC
DeepSeek V4 Flash 0423 StreamLakestreamlake/fp8$0.078 / $0.157 → $0.076 / $0.1512026-08-20 22:05 UTC
DeepSeek V4 Flash 0423 Mancer 2mancer/fp8$0.145 / $0.450 → $0.140 / $0.4502026-08-20 22:00 UTC
DeepSeek V4 Flash 0423 Baidubaidu/fp8$0.080 / $0.160 → $0.077 / $0.1542026-08-20 21:45 UTC
DeepSeek V4 Flash 0423 GMICloudgmicloud/fp8$0.112 / $0.224 → $0.084 / $0.1682026-08-20 18:45 UTC
DeepSeek V4 Flash 0423 StreamLakestreamlake/fp8$0.081 / $0.162 → $0.078 / $0.1572026-08-20 17:55 UTC
DeepSeek V4 Flash 0423 Baidubaidu/fp8$0.083 / $0.165 → $0.080 / $0.1602026-08-20 17:45 UTC
DeepSeek V4 Flash 0423 StreamLakestreamlake/fp8$0.084 / $0.168 → $0.081 / $0.1622026-08-20 16:15 UTC
DeepSeek V4 Flash 0423 Baidubaidu/fp8$0.085 / $0.171 → $0.083 / $0.1652026-08-20 15:50 UTC
DeepSeek V4 Flash 0423 StreamLakestreamlake/fp8$0.087 / $0.174 → $0.084 / $0.1682026-08-20 15:01 UTC
DeepSeek V4 Flash 0423 Baidubaidu/fp8$0.088 / $0.176 → $0.085 / $0.1712026-08-20 14:40 UTC
DeepSeek V4 Flash 0423 StreamLakestreamlake/fp8$0.089 / $0.177 → $0.087 / $0.1742026-08-20 13:01 UTC
DeepSeek V4 Flash 0423 Baidubaidu/fp8$0.140 / $0.280 → $0.088 / $0.1762026-08-20 12:45 UTC
DeepSeek V4 Flash 0423 DeepSeekdeepseek$0.440 / $1.32 → $0.220 / $0.6602026-08-20 10:05 UTC
DeepSeek V4 Flash 0423 DeepSeekdeepseek$0.220 / $0.660 → $0.440 / $1.322026-08-20 06:10 UTC
DeepSeek V4 Flash 0423 DeepSeekdeepseek$0.440 / $1.32 → $0.220 / $0.6602026-08-20 04:05 UTC
DeepSeek V4 Flash 0423 StreamLakestreamlake/fp8$0.083 / $0.165 → $0.089 / $0.1772026-08-20 02:35 UTC
DeepSeek V4 Flash 0423 GMICloudgmicloud/fp8$0.084 / $0.168 → $0.112 / $0.2242026-08-20 01:10 UTC
DeepSeek V4 Flash 0423 DeepSeekdeepseek$0.220 / $0.660 → $0.440 / $1.322026-08-20 01:05 UTC
DeepSeek V4 Flash 0423 StreamLakestreamlake/fp8$0.074 / $0.148 → $0.083 / $0.1652026-08-20 00:00 UTC
DeepSeek V4 Flash 0423 Baidubaidu/fp8$0.076 / $0.151 → $0.140 / $0.2802026-08-19 23:45 UTC
DeepSeek V4 Flash 0423 StreamLakestreamlake/fp8$0.077 / $0.154 → $0.074 / $0.1482026-08-19 22:05 UTC
DeepSeek V4 Flash 0423 Baidubaidu/fp8$0.078 / $0.157 → $0.076 / $0.1512026-08-19 21:50 UTC
DeepSeek V4 Flash 0423 StreamLakestreamlake/fp8$0.080 / $0.160 → $0.077 / $0.1542026-08-19 21:10 UTC
DeepSeek V4 Flash 0423 Baidubaidu/fp8$0.081 / $0.162 → $0.078 / $0.1572026-08-19 20:50 UTC
DeepSeek V4 Flash 0423 StreamLakestreamlake/fp8$0.083 / $0.165 → $0.080 / $0.1602026-08-19 19:15 UTC

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.