DeepSeek V4 Flash

DeepSeek V4 Flash is an efficiency-optimized Mixture-of-Experts model from DeepSeek with 284B total parameters and 13B activated parameters, supporting a 1M-token context window. It is designed for fast inference and high-throughput workloads, while maintaining strong reasoning and coding performance.

The model includes hybrid attention for efficient long-context processing. Reasoning efforts high and xhigh are supported; xhigh maps to max reasoning. It is well suited for applications such as coding assistants, chat systems, and agent workflows where responsiveness and cost efficiency are important.

deepseek/deepseek-v4-flash

Organization: DeepSeek
Family: deepseek-flash
Providers: 53
Context: 1,000,000
Output limit: 384,000
Knowledge: 2025-05
Release: 2026-04-24
Updated: 2026-04-24
Weights: Open
Input: text
Output: text
Capabilities: Tools, Reasoning, Structured, Temperature

Providers

Compare every cataloged offer for this model, including its provider, route, token limits, price, and capabilities.

85 offers · USD per million tokens

Provider	Model ID	Context	Output	Input / output · 1M	Reasoning	Tools	Structured	Details
Abacus	deepseek-ai/DeepSeek-V4-Flash	1,000,000	32,768	$0.14 / $0.28	Yes	Yes	Yes	View
AgentRouter	deepseek-v4-flash	1,000,000	384,000	— / —	Yes	Yes	Yes	View
ai&	deepseek-ai/deepseek-v4-flash	1,048,576	384,000	$0.15 / $0.25	Yes	Yes	Yes	View
Alibaba (China)	deepseek-v4-flash	1,000,000	384,000	$0.14 / $0.28	Yes	Yes	Yes	View
Alibaba Token Plan	deepseek-v4-flash	1,000,000	384,000	$0.00 / $0.00	Yes	Yes	Yes	View
Alibaba Token Plan (China)	deepseek-v4-flash	1,000,000	384,000	$0.00 / $0.00	Yes	Yes	Yes	View
Ambient	deepseek/deepseek-v4-flash	1,048,576	1,048,576	$0.14 / $0.28	Yes	Yes	Yes	View
AnyAPI	deepseek/deepseek-v4-flash	1,000,000	384,000	— / —	Yes	Yes	Yes	View
Auriko	deepseek-v4-flash	1,000,000	384,000	$0.14 / $0.28	Yes	Yes	Yes	View
Azure	deepseek-v4-flash	1,000,000	384,000	$0.19 / $0.51	Yes	No	Yes	View
Charm Hyper	deepseek-v4-flash	1,000,000	384,000	$0.20 / $0.40	Yes	Yes	Yes	View
ClinePass	cline-pass/deepseek-v4-flash	1,000,000	384,000	$0.14 / $0.28	Yes	Yes	Yes	View
CoreWeave	deepseek-ai/DeepSeek-V4-Flash	1,048,576	1,048,576	$0.14 / $0.28	Yes	Yes	Yes	View
CrofAI	deepseek-v4-flash	1,000,000	131,072	$0.12 / $0.21	Yes	Yes	Yes	View
CrossModel	deepseek/deepseek-v4-flash	1,000,000	384,000	$0.30 / $1.20	Yes	Yes	Yes	View
Deep Infra	deepseek-ai/DeepSeek-V4-Flash	1,048,576	16,384	$0.09 / $0.18	Yes	Yes	Yes	View
DevPass (LLM Gateway)	deepseek-v4-flash	1,050,000	384,000	$0.065 / $0.116	Yes	Yes	Yes	View
EBCloud	DeepSeek-V4-Flash	1,000,000	384,000	$0.143 / $0.2857	Yes	Yes	Yes	View
Eden AI	deepseek/deepseek-v4-flash	1,000,000	384,000	$0.15 / $0.60	Yes	Yes	Yes	View
EmpirioLabs AI	deepseek-v4-flash	1,000,000	393,216	$0.14 / $0.28	Yes	Yes	Yes	View
GMI Cloud	deepseek-ai/DeepSeek-V4-Flash	1,048,575	384,000	$0.112 / $0.224	Yes	Yes	Yes	View
HPC-AI	deepseek/deepseek-v4-flash	1,048,576	128,000	$0.14 / $0.28	Yes	Yes	Yes	View
Hugging Face	deepseek-ai/DeepSeek-V4-Flash	1,048,576	384,000	$0.14 / $0.28	Yes	Yes	Yes	View
Impossibl	deepseek/deepseek-v4-flash	1,000,000	384,000	$0.19 / $0.51	Yes	Yes	Yes	View
InferX	deepseek-v4-flash	1,000,000	100,000	$0.00 / $0.00	Yes	Yes	Yes	View
Jalapeno Cloud	DeepSeek-V4-Flash	1,048,576	384,000	$0.14 / $0.28	Yes	Yes	Yes	View
Kenari	deepseek-v4-flash	1,000,000	384,000	$0.00 / $0.00	Yes	Yes	Yes	View
Kenari	deepseek-v4-flash:free	1,000,000	384,000	$0.00 / $0.00	Yes	Yes	Yes	View
Kilo Gateway	deepseek/deepseek-v4-flash	1,048,576	943,718	$0.14 / $0.28	Yes	Yes	Yes	View
LLM Gateway	baidu/deepseek-v4-flash	1,048,576	131,072	$0.44 / $1.32	Yes	Yes	No	View
LLM Gateway	ranoai/deepseek-v4-flash	1,000,000	393,216	$0.14 / $0.28	Yes	Yes	Yes	View
LLM Gateway	consensusprotocol/deepseek-v4-flash	524,288	384,000	$0.10 / $0.20	Yes	Yes	No	View
LLM Gateway	gonka24/deepseek-v4-flash	390,000	16,384	$0.065 / $0.116	Yes	Yes	No	View
LLM Gateway	bytedance/deepseek-v4-flash	1,048,576	393,216	$0.44 / $1.32	Yes	Yes	No	View
LLM Gateway	novita/deepseek-v4-flash	1,050,000	393,216	$0.14 / $0.28	Yes	Yes	No	View
LLM Gateway	tencent/deepseek-v4-flash	1,000,000	393,216	$0.14 / $0.28	Yes	Yes	No	View
LLM Gateway	fireworks/deepseek-v4-flash	1,048,576	393,216	$0.22 / $0.66	Yes	Yes	Yes	View
LLM Gateway	together-ai/deepseek-v4-flash	163,840	163,840	$0.14 / $0.28	Yes	Yes	Yes	View
LLM Gateway	runware/deepseek-v4-flash	1,048,576	384,000	$0.076 / $0.153	Yes	Yes	Yes	View
LLM Gateway	alibaba/deepseek-v4-flash	1,000,000	393,216	$0.20 / $0.40	Yes	Yes	No	View
LLM Gateway	deepinfra/deepseek-v4-flash	1,000,000	393,216	$0.08 / $0.18	Yes	Yes	No	View
LLM Gateway	canopywave/deepseek-v4-flash	1,000,000	393,216	$0.14 / $0.28	Yes	Yes	No	View
Merge Gateway	deepseek/deepseek-v4-flash	1,048,576	384,000	$0.035 / $0.07	Yes	Yes	Yes	View
Modelis	deepseek-v4-flash	1,000,000	384,000	$0.0983 / $0.1966	Yes	Yes	Yes	View
NanoGPT	deepseek/deepseek-v4-flash	1,048,576	384,000	$0.14 / $0.28	Yes	Yes	Yes	View
NanoGPT	deepseek/deepseek-v4-flash:thinking	1,048,576	384,000	$0.14 / $0.28	Yes	Yes	Yes	View
Neuralwatt	deepseek-v4-flash	1,048,560	393,216	$0.14 / $0.28	Yes	Yes	Yes	View
Neuralwatt	deepseek-v4-flash-flex	1,048,560	393,216	$0.091 / $0.182	Yes	Yes	Yes	View
NovitaAI	deepseek/deepseek-v4-flash	1,048,576	393,216	$0.14 / $0.28	Yes	Yes	Yes	View
Nvidia	deepseek-ai/deepseek-v4-flash	1,048,576	393,216	$0.14 / $0.28	Yes	Yes	Yes	View
Ofox	deepseek/deepseek-v4-flash	1,000,000	384,000	$0.44 / $1.32	Yes	Yes	Yes	View
Alibabavia OpenRouter	deepseek/deepseek-v4-flash fp8	1,000,000	393,216	$0.134 / $0.268	Yes	Yes	Yes	View
AtlasCloudvia OpenRouter	deepseek/deepseek-v4-flash fp4	1,048,576	393,216	$0.14 / $0.28	Yes	Yes	Yes	View
Azurevia OpenRouter	deepseek/deepseek-v4-flash unknown	1,048,576	384,000	$0.21 / $0.56	Yes	Yes	No	View
Baiduvia OpenRouter	deepseek/deepseek-v4-flash fp8	1,048,576	131,072	$0.14 / $0.28	Yes	Yes	Yes	View
Cloudflarevia OpenRouter	deepseek/deepseek-v4-flash unknown	384,000	345,600	$0.44 / $1.32	Yes	Yes	Yes	View
DeepInfravia OpenRouter	deepseek/deepseek-v4-flash fp8	1,048,576	65,536	$0.09 / $0.18	Yes	Yes	Yes	View
DigitalOceanvia OpenRouter	deepseek/deepseek-v4-flash unknown	1,048,576	384,000	$0.098 / $0.196	Yes	Yes	Yes	View
GMICloudvia OpenRouter	deepseek/deepseek-v4-flash fp8	1,048,575	943,717	$0.091 / $0.182	Yes	Yes	No	View
Mancer 2via OpenRouter	deepseek/deepseek-v4-flash fp8	1,048,576	943,718	$0.19 / $0.50	Yes	Yes	Yes	View
Novitavia OpenRouter	deepseek/deepseek-v4-flash fp8	1,048,576	393,216	$0.14 / $0.28	Yes	Yes	No	View
OpenInferencevia OpenRouter	deepseek/deepseek-v4-flash fp4	1,048,576	943,718	$0.02912 / $1.664	Yes	Yes	Yes	View
Parasailvia OpenRouter	deepseek/deepseek-v4-flash fp8	1,048,576	943,718	$0.14 / $0.28	Yes	Yes	Yes	View
Relacevia OpenRouter	deepseek/deepseek-v4-flash fp4	1,048,576	943,718	$0.0224 / $1.28	Yes	Yes	No	View
SiliconFlowvia OpenRouter	deepseek/deepseek-v4-flash fp8	1,048,576	393,216	$0.13 / $0.28	Yes	Yes	No	View
StreamLakevia OpenRouter	deepseek/deepseek-v4-flash fp8	1,024,000	384,000	$0.028 / $0.056	Yes	Yes	Yes	View
Venicevia OpenRouter	deepseek/deepseek-v4-flash unknown	1,000,000	32,768	$0.0966 / $0.1925	Yes	Yes	Yes	View
Opper	deepseek-v4-flash	1,000,000	384,000	$0.25 / $0.66	Yes	Yes	Yes	View
OrcaRouter	deepseek/deepseek-v4-flash-free	1,000,000	384,000	$0.00 / $0.00	Yes	Yes	Yes	View
OrcaRouter	deepseek/deepseek-v4-flash	1,000,000	384,000	$0.147 / $0.295	Yes	Yes	Yes	View
Pendra	deepseek-v4-flash	1,000,000	384,000	$0.00 / $0.00	Yes	Yes	Yes	View
Pioneer	deepseek-ai/DeepSeek-V4-Flash	1,000,000	131,072	$0.10 / $0.20	Yes	Yes	Yes	View
Requesty	deepseek-v4-flash	1,048,576	131,072	$0.28 / $0.56	Yes	Yes	Yes	View
routing.run	deepseek-v4-flash	1,000,000	64,000	$0.112 / $0.224	Yes	Yes	Yes	View
SCNet Token Plan	DeepSeek-V4-Flash	1,000,000	384,000	$0.00 / $0.00	Yes	Yes	Yes	View
SenseNova (China)	deepseek-v4-flash	1,000,000	65,536	$0.00 / $0.00	Yes	Yes	Yes	View
SiliconFlow	deepseek-ai/DeepSeek-V4-Flash	1,000,000	384,000	$0.13 / $0.28	Yes	Yes	Yes	View
SiliconFlow (China)	deepseek-ai/DeepSeek-V4-Flash	1,000,000	384,000	$0.14 / $0.28	Yes	Yes	Yes	View
TokenGo	deepseek/deepseek-v4-flash	1,000,000	384,000	$0.098 / $0.196	Yes	Yes	Yes	View
UnoRouter	deepseek-v4-flash:free	1,000,000	384,000	$0.00 / $0.00	Yes	Yes	Yes	View
UnoRouter	deepseek-v4-flash	1,000,000	384,000	$0.0625 / $0.125	Yes	Yes	Yes	View
Venice AI	deepseek-v4-flash	1,000,000	32,768	$0.138 / $0.275	Yes	Yes	Yes	View
Vercel AI Gateway	deepseek/deepseek-v4-flash	1,000,000	384,000	$0.13 / $0.26	Yes	Yes	Yes	View
Volcengine Ark Coding Plan	deepseek-v4-flash	1,000,000	384,000	$0.00 / $0.00	Yes	Yes	Yes	View
ZenMux	deepseek/deepseek-v4-flash	1,000,000	384,000	$0.14 / $0.28	Yes	Yes	Yes	View

Prices and limits apply to each offer. A dash means the value is not provided. OpenRouter offers show the hosting provider and routing channel separately.

Pricing

Compare token rates and cache charges across available routes and hosting configurations.

USD per million tokens

Input from$0.00Per 1M input tokens

Output from$0.00Per 1M output tokens

Available offers85Across 53 providers

Lowest listed input and output prices across the offers above; they may belong to different providers. These are base rates, before conditional pricing, discounts or additional fees.

Endpoint price comparisonLatest observed snapshot · 10 lowest-priced configurations

InputOutput

Relacerelace/fp4

$0.0224$1.28

StreamLakestreamlake/fp8

$0.028$0.056

OpenInferenceopen-inference/fp4

$0.02912$1.664

DeepInfradeepinfra/fp8

$0.09$0.18

GMICloudgmicloud/fp8

$0.091$0.182

Venicevenice

$0.0966$0.1925

DigitalOceandigitalocean

$0.098$0.196

SiliconFlowsiliconflow/fp8

$0.13$0.28

Alibabaalibaba/fp8

$0.134$0.268

AtlasCloudatlas-cloud/fp4

$0.14$0.28

deepseek/deepseek-v4-flashObserved Oct 4, 2026, 11:31 AM UTC

Input: $0.0224
Output: $1.28
Cache read: $0.0224

Hosting prices through OpenRouter All endpoints and cache rates

Endpoint	Input / 1M	Output / 1M	Cache read / 1M	Cache write / 1M
Alibabaalibaba/fp8	$0.134	$0.268	$0.0268	—
AtlasCloudatlas-cloud/fp4	$0.14	$0.28	$0.028	—
Azureazure/us	$0.21	$0.56	$0.031	—
Baidubaidu/fp8	$0.14	$0.28	$0.028	—
Cloudflarecloudflare	$0.44	$1.32	$0.014	—
DeepInfradeepinfra/fp8	$0.09	$0.18	$0.018	—
DigitalOceandigitalocean	$0.098	$0.196	$0.0196	—
GMICloudgmicloud/fp8	$0.091	$0.182	$0.0182	—
Mancer 2mancer/fp8	$0.19	$0.50	—	—
Novitanovita/fp8	$0.14	$0.28	$0.028	—
OpenInferenceopen-inference/fp4	$0.02912	$1.664	$0.02912	—
Parasailparasail/fp8	$0.14	$0.28	$0.07	—
Relacerelace/fp4	$0.0224	$1.28	$0.0224	—
SiliconFlowsiliconflow/fp8	$0.13	$0.28	$0.028	—
StreamLakestreamlake/fp8	$0.028	$0.056	$0.0056	—
Venicevenice	$0.0966	$0.1925	$0.0196	—

Each row is a hosting configuration. Open an offer in Providers for its full pricing conditions and additional fees. Endpoint observations may differ from the routing catalog's base price.

Performance

Review reported latency and output-throughput percentiles from the latest observed rolling window.

Latest 30-minute observation

Time to first token · milliseconds

Latency distributionRolling 30-minute snapshot · 6 endpoints

P50P75P90P99

Alibabaalibaba/fp8

829–5,416.99 ms

AtlasCloudatlas-cloud/fp4

1,665–9,612.6 ms

Azureazure/us

1,128–19,628.32 ms

Baidubaidu/fp8

829–10,612.62 ms

Cloudflarecloudflare

1,371–7,296.82 ms

DeepInfradeepinfra/fp8

750–4,234.42 ms

All latency percentiles 16 endpoints

Endpoint	P50	P75	P90	P99
Alibabaalibaba/fp8	829 ms	946 ms	1,264 ms	5,416.99 ms
AtlasCloudatlas-cloud/fp4	1,665 ms	2,310.25 ms	3,548.2 ms	9,612.6 ms
Azureazure/us	1,128 ms	1,971 ms	4,116.4 ms	19,628.32 ms
Baidubaidu/fp8	829 ms	1,280 ms	2,395.8 ms	10,612.62 ms
Cloudflarecloudflare	1,371 ms	2,211 ms	5,816.4 ms	7,296.82 ms
DeepInfradeepinfra/fp8	750 ms	1,228 ms	1,990.9 ms	4,234.42 ms
DigitalOceandigitalocean	1,772 ms	2,226 ms	2,883.9 ms	16,584.67 ms
GMICloudgmicloud/fp8	1,539 ms	1,773 ms	2,089 ms	4,829.27 ms
Mancer 2mancer/fp8	731 ms	1,129.75 ms	1,810.5 ms	5,935.2 ms
Novitanovita/fp8	1,379 ms	1,708.25 ms	2,377.9 ms	5,109.27 ms
OpenInferenceopen-inference/fp4	836 ms	1,407 ms	2,197.2 ms	6,869.72 ms
Parasailparasail/fp8	753 ms	1,199.25 ms	1,840.9 ms	4,117.36 ms
Relacerelace/fp4	579 ms	998.25 ms	1,685.9 ms	8,558.12 ms
SiliconFlowsiliconflow/fp8	1,154 ms	1,747 ms	2,657.6 ms	7,438.82 ms
StreamLakestreamlake/fp8	1,394.5 ms	1,934 ms	3,082.3 ms	9,169.54 ms
Venicevenice	952 ms	1,333.25 ms	2,021.7 ms	4,838.63 ms

Output throughput · tokens per second

Throughput distributionRolling 30-minute snapshot · 6 endpoints

P50P75P90P99

Alibabaalibaba/fp8

95–151 t/s

AtlasCloudatlas-cloud/fp4

69–119 t/s

Azureazure/us

80.5–218.3 t/s

Baidubaidu/fp8

74–138 t/s

Cloudflarecloudflare

70–84 t/s

DeepInfradeepinfra/fp8

57–113 t/s

All throughput percentiles 16 endpoints

Endpoint	P50	P75	P90	P99
Alibabaalibaba/fp8	95 t/s	114 t/s	130 t/s	151 t/s
AtlasCloudatlas-cloud/fp4	69 t/s	84 t/s	95 t/s	119 t/s
Azureazure/us	80.5 t/s	108 t/s	144 t/s	218.3 t/s
Baidubaidu/fp8	74 t/s	94.25 t/s	110 t/s	138 t/s
Cloudflarecloudflare	70 t/s	77 t/s	80 t/s	84 t/s
DeepInfradeepinfra/fp8	57 t/s	73 t/s	88 t/s	113 t/s
DigitalOceandigitalocean	6 t/s	7 t/s	9 t/s	13 t/s
GMICloudgmicloud/fp8	57 t/s	73 t/s	117 t/s	178 t/s
Mancer 2mancer/fp8	30 t/s	42 t/s	68 t/s	104 t/s
Novitanovita/fp8	63 t/s	86 t/s	108 t/s	174 t/s
OpenInferenceopen-inference/fp4	38 t/s	40 t/s	41 t/s	43 t/s
Parasailparasail/fp8	62 t/s	79 t/s	100 t/s	145.09 t/s
Relacerelace/fp4	84 t/s	117 t/s	160 t/s	285.09 t/s
SiliconFlowsiliconflow/fp8	38 t/s	70 t/s	134 t/s	243.99 t/s
StreamLakestreamlake/fp8	64.5 t/s	83 t/s	98 t/s	114 t/s
Venicevenice	55 t/s	68 t/s	82 t/s	106 t/s

Percentiles describe the observed request distribution. These are snapshots of a rolling window, not a historical time series.

Uptime

Compare successful request rates across the latest 5-minute, 30-minute, and 24-hour observations.

Successful requests by window

Availability by endpointCurrent overlapping windows · 16 endpoints

Endpoint5 minutes30 minutes24 hours

Alibabaalibaba/fp8Observed Oct 4, 2026, 11:31 AM UTC

99.11%

99%

98.98%

AtlasCloudatlas-cloud/fp4Observed Oct 4, 2026, 11:31 AM UTC

99.81%

99.53%

99.11%

Azureazure/usObserved Oct 4, 2026, 11:31 AM UTC

93.36%

96.64%

97.9%

Baidubaidu/fp8Observed Oct 4, 2026, 11:31 AM UTC

99.97%

99.96%

99.2%

CloudflarecloudflareObserved Oct 4, 2026, 11:31 AM UTC

100%

99.95%

DeepInfradeepinfra/fp8Observed Oct 4, 2026, 11:31 AM UTC

99.97%

99.98%

99.88%

DigitalOceandigitaloceanObserved Oct 4, 2026, 11:31 AM UTC

99.98%

99.93%

99.89%

GMICloudgmicloud/fp8Observed Oct 4, 2026, 11:31 AM UTC

100%

99.99%

Mancer 2mancer/fp8Observed Oct 4, 2026, 11:31 AM UTC

98.55%

99.39%

98.2%

Novitanovita/fp8Observed Oct 4, 2026, 11:31 AM UTC

99.95%

99.93%

99.89%

OpenInferenceopen-inference/fp4Observed Oct 4, 2026, 11:31 AM UTC

99.87%

99.86%

99.42%

Parasailparasail/fp8Observed Oct 4, 2026, 11:31 AM UTC

100%

99.93%

99.81%

Relacerelace/fp4Observed Oct 4, 2026, 11:31 AM UTC

99.94%

99.89%

99.87%

SiliconFlowsiliconflow/fp8Observed Oct 4, 2026, 11:31 AM UTC

99.76%

99.97%

99.66%

StreamLakestreamlake/fp8Observed Oct 4, 2026, 11:31 AM UTC

99.86%

99.87%

99.45%

VeniceveniceObserved Oct 4, 2026, 11:31 AM UTC

99.52%

99.55%

99.25%

Reported by OpenRouter at the time of collection; rate-limited requests are excluded. A dash means no measurement was reported. This is not a live availability check.

Benchmarks

Review attributed evaluation results on each source's original scale.

External evaluations

deepseek/deepseek-v4-flashObserved Oct 4, 2026, 11:31 AM UTC

Benchmark profileArtificial Analysis · shared source scale from 0 to 60

Coding

Agentic

26.3

Intelligence

24.4

Design Arena	Category	Elo	Win rate	Rank
models	3d	1,204	49.3%	46
models	asciiart	1,122	42.8%	55
models	codecategories	1,211	48.9%	49
models	dataviz	1,136	40.7%	83
models	gamedev	1,210	50.2%	46
models	svg	1,155	48.4%	40
models	uicomponent	1,171	44.7%	66
models	website	1,214	49.1%	51

Sources: Artificial Analysis and Design Arena, via OpenRouter. Scores retain their original scales and are not OpenAI Suite ratings.

Detailed evaluations

Evaluation scoresAccuracy or item F1 · higher is better

gpqa diamondDeepSeek: DeepSeek V4 Flash 0423

85.831%

tau bench verified airlineDeepSeek: DeepSeek V4 Flash 0423

74.444%

Evaluation	Configuration	Source	Results	Details
models · dataviz	DeepSeek-V4-Flashdeepseek/deepseek-v4-flash-20260423	Design Arena	1,136 Elo	Details
models · uicomponent	DeepSeek-V4-Flashdeepseek/deepseek-v4-flash-20260423	Design Arena	1,171 Elo	Details
gpqa diamond	DeepSeek: DeepSeek V4 Flash 0423deepseek/deepseek-v4-flash-20260423	OpenRouter	85.831%	Details
models · asciiart	DeepSeek-V4-Flashdeepseek/deepseek-v4-flash-20260423	Design Arena	1,122 Elo	Details
Composite indices	DeepSeek V4 Flash 0420 (Reasoning, High Effort)deepseek/deepseek-v4-flash-20260423	Artificial Analysis	Intelligence 24.4 · Coding 52 · Agentic 26.3	Details
models · website	DeepSeek-V4-Flashdeepseek/deepseek-v4-flash-20260423	Design Arena	1,214 Elo	Details
models · codecategories	DeepSeek-V4-Flashdeepseek/deepseek-v4-flash-20260423	Design Arena	1,211 Elo	Details
tau bench verified airline	DeepSeek: DeepSeek V4 Flash 0423deepseek/deepseek-v4-flash-20260423	OpenRouter	74.444%	Details
models · svg	DeepSeek-V4-Flashdeepseek/deepseek-v4-flash-20260423	Design Arena	1,155 Elo	Details
models · 3d	DeepSeek-V4-Flashdeepseek/deepseek-v4-flash-20260423	Design Arena	1,204 Elo	Details
models · gamedev	DeepSeek-V4-Flashdeepseek/deepseek-v4-flash-20260423	Design Arena	1,210 Elo	Details

Sources: Artificial Analysis, Design Arena and OpenRouter. Evaluation configurations and scales differ; results are not interchangeable. Updated Oct 4, 2026.

Apps & session costs

Compare observed median session costs for applications using this model, grouped by conversation length.

30-day sample · through Sep 27, 2026

Application	Model route	Turns	Median cost / session	Details
Claude Code	deepseek/deepseek-v4-flash-20260423	50+ turns	$0.340484	Details
Hermes Agent	deepseek/deepseek-v4-flash-20260423	1 turn	$0.001051	Details
Kilo Code	deepseek/deepseek-v4-flash-20260423	1 turn	$0.000242	Details
Kilo Code	deepseek/deepseek-v4-flash-20260423	50+ turns	$0.231597	Details
Claude Code	deepseek/deepseek-v4-flash-20260423	2–9 turns	$0.004733	Details
Kilo Code	deepseek/deepseek-v4-flash-20260423	2–9 turns	$0.005149	Details
Hermes Agent	deepseek/deepseek-v4-flash-20260423	2–9 turns	$0.003956	Details
Hermes Agent	deepseek/deepseek-v4-flash-20260423	50+ turns	$0.1802	Details
Claude Code	deepseek/deepseek-v4-flash-20260423	10–49 turns	$0.035865	Details
Claude Code	deepseek/deepseek-v4-flash-20260423	1 turn	$0.001007	Details
Hermes Agent	deepseek/deepseek-v4-flash-20260423	10–49 turns	$0.028423	Details
Kilo Code	deepseek/deepseek-v4-flash-20260423	10–49 turns	$0.026786	Details
Codex	deepseek/deepseek-v4-flash-20260423	1 turn	$0.000479	Details

Source: OpenRouter, as of Oct 2, 2026. Published under CC BY 4.0. These are observed costs per session, not token prices or a forecast for your workload. Apps are compared separately.

Activity

Daily prompt and completion tokens reported for this model's routes in OpenRouter's top 50. Missing days have no published value.

Last 90 complete UTC days

Daily token usage64,340,496,361,578 reported tokens · 89 days with observations

1.2T0

2026-07-07
806,188,922,027 tokens

2026-07-08
839,437,484,340 tokens

2026-07-09
758,577,549,228 tokens

2026-07-10
734,964,582,490 tokens

2026-07-11
706,863,897,979 tokens

2026-07-12
628,873,101,299 tokens

2026-07-13
763,497,210,011 tokens

2026-07-14
823,161,262,273 tokens

2026-07-15
805,869,355,388 tokens

2026-07-16
777,169,564,648 tokens

2026-07-17
804,551,330,089 tokens

2026-07-18
738,407,163,994 tokens

2026-07-19
672,296,882,701 tokens

2026-07-20
713,920,960,951 tokens

2026-07-21
857,711,413,156 tokens

2026-07-22
913,235,097,630 tokens

2026-07-23
888,638,863,216 tokens

2026-07-24
1,024,483,223,873 tokens

2026-07-25
943,926,033,735 tokens

2026-07-26
1,028,767,203,083 tokens

2026-07-27
1,074,240,787,259 tokens

2026-07-28
1,126,445,711,067 tokens

2026-07-29
1,210,912,290,867 tokens

2026-07-30
1,161,553,833,401 tokens

2026-07-31
1,014,960,283,911 tokens

2026-08-01
814,919,754,746 tokens

2026-08-02
817,317,462,750 tokens

2026-08-03
950,747,011,708 tokens

2026-08-04
949,465,311,824 tokens

2026-08-05
890,359,448,879 tokens

2026-08-06
869,720,046,288 tokens

2026-08-07
856,924,390,802 tokens

2026-08-08
715,116,334,631 tokens

2026-08-09
648,100,578,381 tokens

2026-08-10
787,395,180,090 tokens

2026-08-11
730,328,820,072 tokens

2026-08-12
688,321,164,241 tokens

2026-08-13
684,590,392,020 tokens

2026-08-14
645,324,891,208 tokens

2026-08-15
581,065,793,450 tokens

2026-08-16
713,184,199,963 tokens

2026-08-17
730,723,224,283 tokens

2026-08-18
738,060,687,036 tokens

2026-08-19
790,371,615,212 tokens

2026-08-20
861,314,777,146 tokens

2026-08-21
782,155,149,557 tokens

2026-08-22
681,172,064,573 tokens

2026-08-23
874,995,733,113 tokens

2026-08-24
805,514,972,154 tokens

2026-08-25
781,657,288,128 tokens

2026-08-26
782,467,612,078 tokens

2026-08-27
784,785,952,575 tokens

2026-08-28
711,480,301,714 tokens

2026-08-29
663,247,324,200 tokens

2026-08-30
674,321,451,836 tokens

2026-08-31
753,776,079,392 tokens

2026-09-01
792,305,220,829 tokens

2026-09-02
798,327,433,150 tokens

2026-09-03
787,058,518,989 tokens

2026-09-04
760,911,072,222 tokens

2026-09-05
668,913,736,680 tokens

2026-09-06
626,624,764,221 tokens

2026-09-07
675,694,267,638 tokens

2026-09-08
675,685,390,547 tokens

2026-09-09
645,268,517,224 tokens

2026-09-10
703,354,448,295 tokens

2026-09-11
632,142,353,026 tokens

2026-09-12
506,337,684,991 tokens

2026-09-13
519,252,847,480 tokens

2026-09-14
647,696,862,304 tokens

2026-09-15
668,011,906,315 tokens

2026-09-16
564,341,684,691 tokens

2026-09-17
555,784,110,039 tokens

2026-09-18
491,306,144,680 tokens

2026-09-19
427,456,307,998 tokens

2026-09-20
418,200,633,146 tokens

2026-09-21
529,315,316,586 tokens

2026-09-22
523,702,324,427 tokens

2026-09-23
514,836,984,021 tokens

2026-09-24
503,033,489,694 tokens

2026-09-25
465,135,278,327 tokens

2026-09-26
412,391,476,621 tokens

2026-09-27
402,429,560,914 tokens

2026-09-28
470,246,427,688 tokens

2026-09-29
478,962,779,354 tokens

2026-09-30
472,610,525,257 tokens

2026-10-01
505,024,762,291 tokens

2026-10-02
522,266,242,527 tokens

2026-10-03
398,288,266,740 tokens

Jul 6, 2026Oct 3, 2026

Daily observations · 89 days

Date (UTC)	Reported tokens	Reported routes
Oct 3, 2026	398,288,266,740	1
Oct 2, 2026	522,266,242,527	1
Oct 1, 2026	505,024,762,291	1
Sep 30, 2026	472,610,525,257	1
Sep 29, 2026	478,962,779,354	1
Sep 28, 2026	470,246,427,688	1
Sep 27, 2026	402,429,560,914	1
Sep 26, 2026	412,391,476,621	1
Sep 25, 2026	465,135,278,327	1
Sep 24, 2026	503,033,489,694	1
Sep 23, 2026	514,836,984,021	1
Sep 22, 2026	523,702,324,427	1
Sep 21, 2026	529,315,316,586	1
Sep 20, 2026	418,200,633,146	1
Sep 19, 2026	427,456,307,998	1
Sep 18, 2026	491,306,144,680	1
Sep 17, 2026	555,784,110,039	1
Sep 16, 2026	564,341,684,691	1
Sep 15, 2026	668,011,906,315	1
Sep 14, 2026	647,696,862,304	1
Sep 13, 2026	519,252,847,480	1
Sep 12, 2026	506,337,684,991	1
Sep 11, 2026	632,142,353,026	1
Sep 10, 2026	703,354,448,295	1
Sep 9, 2026	645,268,517,224	1
Sep 8, 2026	675,685,390,547	1
Sep 7, 2026	675,694,267,638	1
Sep 6, 2026	626,624,764,221	1
Sep 5, 2026	668,913,736,680	1
Sep 4, 2026	760,911,072,222	1
Sep 3, 2026	787,058,518,989	1
Sep 2, 2026	798,327,433,150	1
Sep 1, 2026	792,305,220,829	1
Aug 31, 2026	753,776,079,392	1
Aug 30, 2026	674,321,451,836	1
Aug 29, 2026	663,247,324,200	1
Aug 28, 2026	711,480,301,714	1
Aug 27, 2026	784,785,952,575	1
Aug 26, 2026	782,467,612,078	1
Aug 25, 2026	781,657,288,128	1
Aug 24, 2026	805,514,972,154	1
Aug 23, 2026	874,995,733,113	1
Aug 22, 2026	681,172,064,573	1
Aug 21, 2026	782,155,149,557	1
Aug 20, 2026	861,314,777,146	1
Aug 19, 2026	790,371,615,212	1
Aug 18, 2026	738,060,687,036	1
Aug 17, 2026	730,723,224,283	1
Aug 16, 2026	713,184,199,963	1
Aug 15, 2026	581,065,793,450	1
Aug 14, 2026	645,324,891,208	1
Aug 13, 2026	684,590,392,020	1
Aug 12, 2026	688,321,164,241	1
Aug 11, 2026	730,328,820,072	1
Aug 10, 2026	787,395,180,090	1
Aug 9, 2026	648,100,578,381	1
Aug 8, 2026	715,116,334,631	1
Aug 7, 2026	856,924,390,802	1
Aug 6, 2026	869,720,046,288	1
Aug 5, 2026	890,359,448,879	1
Aug 4, 2026	949,465,311,824	1
Aug 3, 2026	950,747,011,708	1
Aug 2, 2026	817,317,462,750	1
Aug 1, 2026	814,919,754,746	1
Jul 31, 2026	1,014,960,283,911	1
Jul 30, 2026	1,161,553,833,401	1
Jul 29, 2026	1,210,912,290,867	1
Jul 28, 2026	1,126,445,711,067	1
Jul 27, 2026	1,074,240,787,259	1
Jul 26, 2026	1,028,767,203,083	1
Jul 25, 2026	943,926,033,735	1
Jul 24, 2026	1,024,483,223,873	1
Jul 23, 2026	888,638,863,216	1
Jul 22, 2026	913,235,097,630	1
Jul 21, 2026	857,711,413,156	1
Jul 20, 2026	713,920,960,951	1
Jul 19, 2026	672,296,882,701	1
Jul 18, 2026	738,407,163,994	1
Jul 17, 2026	804,551,330,089	1
Jul 16, 2026	777,169,564,648	1
Jul 15, 2026	805,869,355,388	1
Jul 14, 2026	823,161,262,273	1
Jul 13, 2026	763,497,210,011	1
Jul 12, 2026	628,873,101,299	1
Jul 11, 2026	706,863,897,979	1
Jul 10, 2026	734,964,582,490	1
Jul 9, 2026	758,577,549,228	1
Jul 8, 2026	839,437,484,340	1
Jul 7, 2026	806,188,922,027	1

Source: OpenRouter (openrouter.ai/rankings), as of Oct 4, 2026. CC BY 4.0. Totals include only individually published routes. Tokenizers differ by provider; the aggregated “other” category is never assigned to a model.

Technical details & API

Explore specifications, supported parameters, reasoning controls, and API identifiers for each OpenRouter route.

Current route specifications

deepseek/deepseek-v4-flashObserved Oct 4, 2026, 11:31 AM UTC

Context: 1,048,576
Maximum output: 943,718
Tokenizer: DeepSeek
Input: text
Output: text
Moderation: No
Added to OpenRouter: 2026-04-24

Base URLhttps://openrouter.ai/api/v1

Model IDdeepseek/deepseek-v4-flash

Version IDdeepseek/deepseek-v4-flash-20260423

Weights IDdeepseek-ai/DeepSeek-V4-Flash

Reasoning

Required: No
Enabled by default: Not reported
Default effort: high
Supported efforts: xhigh, high

Supported parameters

frequency_penalty
include_reasoning
logit_bias
logprobs
max_completion_tokens
max_tokens
min_p
presence_penalty
reasoning
reasoning_effort
repetition_penalty
response_format
seed
stop
structured_outputs
temperature
tool_choice
tools
top_a
top_k
top_logprobs
top_p

Frequently asked questions

Answers based on the model metadata and serving offers currently in the catalog.

How much does DeepSeek V4 Flash cost?

Listed rates start at $0.00 per million input tokens and $0.00 per million output tokens. Prices vary by provider and configuration; see Pricing for details.

Which providers offer DeepSeek V4 Flash?

There are 85 cataloged offers across 53 providers. Compare model IDs, prices and limits in Providers.

What is the context window?

The cataloged model context is 1,000,000 tokens. Each serving endpoint may apply a different limit.

Does it support tools and structured output?

Tool calling: Yes. Structured output: Yes. Support can vary by endpoint.

Resources

Hugging Face ↗

Catalog history · 2 recorded revisions

Metadata observations since this model was first cataloged. These are not model release versions.

Sep 14, 2026 · 3dbdcfc38310
Sep 12, 2026 · 3cd8639e68e4