DeepSeek V4.1 Flash

DeepSeek V4.1 Flash is a sparse mixture-of-experts model from DeepSeek, and the first built on the company's Causal Encoder-Decoder (CED) architecture. It activates 8B parameters on input and 16B on output from a 552B-parameter backbone, an asymmetric split that keeps per-token compute low relative to the model's total size. Image understanding is native to the architecture, with visual and text embeddings trained jointly from the start of pre-training rather than added afterward as in the earlier experimental V4 Flash Vision Exp.

It is suited for coding, terminal, and computer-use agents, along with long-horizon tasks that must run to completion across many steps and long-context analysis. Compressed KV caching cuts cache memory to roughly a quarter of the previous Flash generation, significantly reducing costs on agentic workloads. DeepSeek positions it as the cost-efficient tier of the V4.1 family and reports that it exceeds V4 Pro on performance, speed, and task completion time.

deepseek/deepseek-v4.1-flash

Organization: DeepSeek
Family: deepseek-flash
Providers: 68
Context: 1,000,000
Output limit: 384,000
Knowledge: 2025-05
Release: 2026-09-10
Updated: —
Weights: Open
Input: text, image/jpeg, image/png, image/gif, image/webp
Output: text
Capabilities: —

Providers

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

129 offers · USD per million tokens

Provider	Model ID	Context	Output	Input / output · 1M	Reasoning	Tools	Structured	Details
302.AI	deepseek-flash	1,000,000	384,000	— / —	—	—	—	View
above.dev	deepseek-v4-flash	1,000,000	384,000	— / —	—	—	—	View
above.dev	deepseek-v4.1-flash	1,000,000	384,000	— / —	—	—	—	View
ai&	deepseek-ai/deepseek-v4.1-flash	1,048,576	384,000	— / —	—	—	—	View
AIHubMix	deepseek-v4.1-flash	1,000,000	384,000	— / —	—	—	—	View
AI/ML API	deepseek/deepseek-v4.1-flash	1,048,576	384,000	— / —	—	—	—	View
AKI.IO	deepseek-v4.1-flash-552b	1,048,576	81,920	— / —	—	—	—	View
Alibaba	deepseek-v4.1-flash	1,000,000	384,000	— / —	—	—	—	View
Alibaba (China)	deepseek-v4.1-flash	1,000,000	384,000	— / —	—	—	—	View
Alibaba Token Plan	deepseek-v4.1-flash	1,000,000	384,000	— / —	—	—	—	View
Alibaba Token Plan (China)	deepseek-v4.1-flash	1,000,000	384,000	— / —	—	—	—	View
Arcee	deepseek/deepseek-v4.1-flash	1,000,000	384,000	— / —	—	—	—	View
Azure	deepseek-v4.1-flash	1,000,000	128,000	— / —	—	—	—	View
Baseten	deepseek-ai/DeepSeek-V4.1-Flash-Fast	1,048,576	32,768	— / —	—	—	—	View
Baseten	deepseek-ai/DeepSeek-V4.1-Flash	1,048,576	32,768	— / —	—	—	—	View
Boundless	deepseek-v4.1-flash	1,048,576	384,000	— / —	—	—	—	View
Charm Hyper	deepseek-v4.1-flash	1,048,576	262,144	— / —	—	—	—	View
ClinePass	cline-pass/deepseek-v4.1-flash	1,000,000	384,000	— / —	—	—	—	View
CommandCode	deepseek/deepseek-v4.1-flash	1,000,000	384,000	— / —	—	—	—	View
CommandCode	deepseek/deepseek-v4.1-flash-fast	1,000,000	384,000	— / —	—	—	—	View
CoralBricks	deepseek-v4.1-flash-fast	1,048,576	393,216	— / —	—	—	—	View
CoreWeave	deepseek-ai/DeepSeek-V4.1-Flash	1,048,576	1,048,576	— / —	—	—	—	View
Cortecs	deepseek-v4.1-flash	1,048,576	393,216	— / —	—	—	—	View
CrossModel	deepseek/deepseek-v4.1-flash	1,000,000	384,000	— / —	—	—	—	View
Deep Infra	deepseek-ai/DeepSeek-V4.1-Flash	1,048,576	384,000	— / —	—	—	—	View
DevPass (LLM Gateway)	deepseek-v4.1-flash	1,050,000	384,000	— / —	—	—	—	View
DigitalOcean	deepseek-v4.1-flash	1,048,576	1,048,576	— / —	—	—	—	View
Eden AI	flexai/DeepSeek-V4.1-Flash	1,048,576	384,000	— / —	—	—	—	View
Eden AI	greenference/deepseek-v4.1-flash	1,048,576	384,000	— / —	—	—	—	View
Eden AI	greenpt/deepseek-v4.1-flash	1,000,000	384,000	— / —	—	—	—	View
Eden AI	tensorx/deepseek/deepseek-v4.1-flash	1,048,576	384,000	— / —	—	—	—	View
Eden AI	deepinfra/deepseek-ai/DeepSeek-V4.1-Flash	1,048,576	384,000	— / —	—	—	—	View
Eden AI	arkanecloud/deepseek-ai/DeepSeek-V4.1-Flash	1,048,576	384,000	— / —	—	—	—	View
Eden AI	qwen/deepseek-v4.1-flash	1,000,000	384,000	— / —	—	—	—	View
Eden AI	nebius/deepseek-ai/DeepSeek-V4.1-Flash	1,048,000	384,000	— / —	—	—	—	View
Eden AI	together_ai/deepseek-ai/DeepSeek-V4.1-Flash	1,048,576	384,000	— / —	—	—	—	View
Eden AI	tencent/deepseek-v4.1-flash	1,000,000	384,000	— / —	—	—	—	View
Eden AI	scaleway/deepseek-v4.1-flash	256,000	384,000	— / —	—	—	—	View
EmpirioLabs AI	deepseek-v4-1-flash	1,000,000	393,216	— / —	—	—	—	View
engy	deepseek-v4.1-flash	1,048,576	65,536	— / —	—	—	—	View
Fireworks AI	accounts/fireworks/routers/deepseek-v4p1-flash-spot	1,000,000	384,000	— / —	—	—	—	View
Fireworks AI	accounts/fireworks/models/deepseek-v4p1-flash	1,000,000	384,000	— / —	—	—	—	View
GreenPT	deepseek-v4.1-flash	1,000,000	384,000	— / —	—	—	—	View
Hugging Face	deepseek-ai/DeepSeek-V4.1-Flash	1,048,576	384,000	— / —	—	—	—	View
Inco	deepseek-v4.1-flash:fast	1,000,000	384,000	— / —	—	—	—	View
IteraCompute	deepseek/deepseek-v4.1-flash	1,048,576	393,216	— / —	—	—	—	View
IteraCompute	deepseek/deepseek-v4-flash-0731	1,048,576	393,216	— / —	—	—	—	View
Kenari	deepseek-v4-1-flash	1,000,000	384,000	— / —	—	—	—	View
Kilo Gateway	deepseek/deepseek-v4.1-flash	1,048,576	943,718	— / —	—	—	—	View
LLM Gateway	novita/deepseek-v4.1-flash	1,048,576	393,216	— / —	—	—	—	View
LLM Gateway	scx-ai-gp/deepseek-v4.1-flash	1,000,000	384,000	— / —	—	—	—	View
LLM Gateway	baidu/deepseek-v4.1-flash	1,048,576	393,216	— / —	—	—	—	View
LLM Gateway	reka-ai/deepseek-v4.1-flash	1,048,576	393,216	— / —	—	—	—	View
LLM Gateway	together-ai/deepseek-v4.1-flash	1,048,576	393,216	— / —	—	—	—	View
LLM Gateway	alibaba/deepseek-v4.1-flash	1,000,000	393,216	— / —	—	—	—	View
LLM Gateway	luminal/deepseek-v4.1-flash	1,000,000	384,000	— / —	—	—	—	View
LLM Gateway	runware/deepseek-v4.1-flash	1,048,576	384,000	— / —	—	—	—	View
LLM Gateway	deepinfra/deepseek-v4.1-flash	1,048,576	393,216	— / —	—	—	—	View
LLM Gateway	consensusprotocol/deepseek-v4.1-flash	1,048,576	384,000	— / —	—	—	—	View
LLM Gateway	fireworks/deepseek-v4.1-flash	1,048,576	393,216	— / —	—	—	—	View
LLM Gateway	entrim/deepseek-v4.1-flash	1,048,576	384,000	— / —	—	—	—	View
Melious	deepseek-v4.1-flash	1,000,000	384,000	— / —	—	—	—	View
Merge Gateway	deepseek/deepseek-v4.1-flash	1,000,000	384,000	— / —	—	—	—	View
Modal	deepseek-ai/DeepSeek-V4.1-Flash	1,000,000	384,000	— / —	—	—	—	View
NaN	deepseek-v4-flash	1,000,000	384,000	— / —	—	—	—	View
NanoGPT	deepseek/deepseek-v4.1-flash	1,000,000	384,000	— / —	—	—	—	View
NanoGPT	TEE/deepseek-v4.1-flash	1,048,576	384,000	— / —	—	—	—	View
NanoGPT	deepseek/deepseek-v4.1-flash:thinking	1,000,000	384,000	— / —	—	—	—	View
Nebius Token Factory	deepseek-ai/DeepSeek-V4.1-Flash	1,048,000	1,048,000	— / —	—	—	—	View
Neon	deepseek-v4-1-flash	1,000,000	384,000	— / —	—	—	—	View
Novita AI	deepseek/deepseek-v4.1-flash	1,048,576	393,216	— / —	—	—	—	View
Nvidia	deepseek-ai/deepseek-v4.1-flash	1,000,000	384,000	— / —	—	—	—	View
Ofox	deepseek/deepseek-v4.1-flash	1,000,000	384,000	— / —	—	—	—	View
Ollama Cloud	deepseek-v4.1-flash	1,048,576	384,000	— / —	—	—	—	View
OpenCode	deepseek-v4.1-flash-free	200,000	32,000	— / —	—	—	—	View
OpenCode	deepseek-v4.1-flash	1,000,000	384,000	— / —	—	—	—	View
OpenCode Go	deepseek-v4.1-flash	1,000,000	384,000	— / —	—	—	—	View
Alibabavia OpenRouter	deepseek/deepseek-v4.1-flash unknown	1,000,000	393,216	$0.30 / $1.20	Yes	Yes	No	View
AtlasCloudvia OpenRouter	deepseek/deepseek-v4.1-flash fp8	1,048,576	393,216	$0.30 / $1.20	Yes	Yes	Yes	View
Baiduvia OpenRouter	deepseek/deepseek-v4.1-flash fp8	1,048,576	393,216	$0.30 / $1.20	Yes	Yes	Yes	View
BaseTenvia OpenRouter	deepseek/deepseek-v4.1-flash fp32	1,048,576	131,072	$0.60 / $2.40	Yes	No	No	View
BaseTenvia OpenRouter	deepseek/deepseek-v4.1-flash fp8	1,048,576	262,144	$0.30 / $1.20	Yes	Yes	Yes	View
CoreWeavevia OpenRouter	deepseek/deepseek-v4.1-flash fp8	1,048,576	943,718	$0.20 / $0.65	Yes	Yes	Yes	View
Crusoevia OpenRouter	deepseek/deepseek-v4.1-flash fp8	1,048,576	943,718	$0.29 / $1.20	Yes	Yes	Yes	View
Decartvia OpenRouter	deepseek/deepseek-v4.1-flash fp4	1,048,576	262,144	$0.09 / $0.18	Yes	Yes	Yes	View
DeepInfravia OpenRouter	deepseek/deepseek-v4.1-flash fp8	1,048,576	131,072	$0.14 / $0.42	Yes	Yes	Yes	View
DeepSeekvia OpenRouter	deepseek/deepseek-v4.1-flash unknown	1,048,576	393,216	$0.15 / $0.60	Yes	Yes	No	View
DekaLLMvia OpenRouter	deepseek/deepseek-v4.1-flash unknown	1,048,576	943,718	$0.12 / $1.20	Yes	Yes	Yes	View
DigitalOceanvia OpenRouter	deepseek/deepseek-v4.1-flash unknown	1,048,576	943,718	$0.165 / $0.66	Yes	Yes	Yes	View
Fireworksvia OpenRouter	deepseek/deepseek-v4.1-flash unknown	1,048,576	943,718	$0.30 / $1.20	Yes	Yes	Yes	View
Fireworksvia OpenRouter	deepseek/deepseek-v4.1-flash unknown	1,048,576	943,718	$0.45 / $1.80	Yes	Yes	Yes	View
GMICloudvia OpenRouter	deepseek/deepseek-v4.1-flash fp8	1,048,575	943,717	$0.18 / $0.72	Yes	Yes	No	View
InferenceNetvia OpenRouter	deepseek/deepseek-v4.1-flash fp8	1,040,000	393,216	$0.07 / $0.60	Yes	Yes	Yes	View
Ionstreamvia OpenRouter	deepseek/deepseek-v4.1-flash unknown	1,048,576	943,718	$0.10 / $1.10	Yes	Yes	Yes	View
Makoravia OpenRouter	deepseek/deepseek-v4.1-flash fp8	1,048,576	393,216	$0.27 / $1.15	Yes	Yes	Yes	View
Modalvia OpenRouter	deepseek/deepseek-v4.1-flash unknown	1,048,576	943,718	$0.30 / $1.20	Yes	Yes	Yes	View
Morphvia OpenRouter	deepseek/deepseek-v4.1-flash fp8	1,048,576	943,718	$0.029 / $1.00	Yes	Yes	Yes	View
Novitavia OpenRouter	deepseek/deepseek-v4.1-flash fp8	1,048,576	393,216	$0.195 / $0.78	Yes	Yes	No	View
OpenInferencevia OpenRouter	deepseek/deepseek-v4.1-flash fp4	1,048,576	943,718	$0.00011 / $0.36	Yes	Yes	Yes	View
Parasailvia OpenRouter	deepseek/deepseek-v4.1-flash fp8	1,048,576	943,718	$0.30 / $1.20	Yes	Yes	Yes	View
Phalavia OpenRouter	deepseek/deepseek-v4.1-flash unknown	1,048,576	943,718	$0.21 / $0.84	Yes	Yes	No	View
Relacevia OpenRouter	deepseek/deepseek-v4.1-flash unknown	1,048,576	943,718	$0.0001 / $0.60	Yes	Yes	No	View
Sail Researchvia OpenRouter	deepseek/deepseek-v4.1-flash fp4	1,048,576	384,000	$0.08 / $0.40	Yes	Yes	Yes	View
SiliconFlowvia OpenRouter	deepseek/deepseek-v4.1-flash fp8	1,048,576	393,216	$0.30 / $1.20	Yes	Yes	No	View
StreamLakevia OpenRouter	deepseek/deepseek-v4.1-flash fp8	1,024,000	384,000	$0.147 / $0.588	Yes	Yes	No	View
Togethervia OpenRouter	deepseek/deepseek-v4.1-flash unknown	1,048,576	943,718	$0.30 / $1.20	Yes	Yes	Yes	View
Venicevia OpenRouter	deepseek/deepseek-v4.1-flash fp8	1,000,000	131,072	$0.30 / $1.20	Yes	Yes	Yes	View
Wafervia OpenRouter	deepseek/deepseek-v4.1-flash unknown	1,048,576	943,718	$0.05 / $1.60	Yes	Yes	Yes	View
Prime Inference	deepseek/deepseek-v4.1-flash	1,048,576	131,072	— / —	—	—	—	View
Requesty	deepseek-v4.1-flash	1,048,576	393,216	— / —	—	—	—	View
Requesty	deepseek-v4.1-flash@eu	1,048,576	393,216	— / —	—	—	—	View
Scaleway	deepseek-v4.1-flash	256,000	32,768	— / —	—	—	—	View
SCNet Token Plan	DeepSeek-V4.1-Flash	1,000,000	384,000	— / —	—	—	—	View
SenseNova (China)	deepseek-v4.1-flash	1,048,576	65,536	— / —	—	—	—	View
Standard Compute	deepseek/deepseek-v4.1-flash	1,000,000	50,000	— / —	—	—	—	View
Subconscious	subconscious/deepseek-v4.1-flash-marathon	5,000,000	384,000	— / —	—	—	—	View
Synthetic	hf:deepseek-ai/DeepSeek-V4.1-Flash	524,288	65,536	— / —	—	—	—	View
Tempr Gateway	deepseek/deepseek-flash	1,000,000	384,000	— / —	—	—	—	View
TensorX	deepseek/deepseek-v4.1-flash	1,048,576	384,000	— / —	—	—	—	View
Thinking Machines	deepseek-ai/DeepSeek-V4.1-Flash:peft:262144	262,144	262,144	— / —	—	—	—	View
Tinfoil	deepseek-v4-1-flash	1,048,576	384,000	— / —	—	—	—	View
Together AI	deepseek-ai/DeepSeek-V4.1-Flash	1,048,576	384,000	— / —	—	—	—	View
Vancine	deepseek-v4.1-flash	1,000,000	384,000	— / —	—	—	—	View
Venice AI	deepseek-v4-1-flash	1,000,000	131,072	— / —	—	—	—	View
Vercel AI Gateway	deepseek/deepseek-v4.1-flash-fast	1,048,576	32,768	— / —	—	—	—	View
Vercel AI Gateway	deepseek/deepseek-v4.1-flash	1,048,576	32,768	— / —	—	—	—	View
Vivgrid	deepseek-v4.1-flash	1,000,000	384,000	— / —	—	—	—	View
Vultr	deepseek-v4.1-flash	1,048,576	384,000	— / —	—	—	—	View
ZenMux	deepseek/deepseek-v4.1-flash	1,000,000	384,000	— / —	—	—	—	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$0.0001Per 1M input tokens

Output from$0.18Per 1M output tokens

Available offers129Across 68 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

$0.0001$0.60

OpenInferenceopen-inference/fp4

$0.00011$0.36

Morphmorph/fp8

$0.029$1.00

Waferwafer

$0.05$1.60

InferenceNetinference-net/fp8

$0.07$0.60

Sail Researchsail-research/fp4

$0.08$0.40

Decartdecart/fp4

$0.09$0.18

Ionstreamionstream

$0.10$1.10

DekaLLMdekallm

$0.12$1.20

DeepInfradeepinfra/fp8

$0.14$0.42

deepseek/deepseek-v4.1-flashObserved Oct 10, 2026, 11:31 AM UTC

Input: $0.30
Output: $1.20
Cache read: $0.006

Hosting prices through OpenRouter All endpoints and cache rates

Endpoint	Input / 1M	Output / 1M	Cache read / 1M	Cache write / 1M
Alibabaalibaba	$0.30	$1.20	$0.03	—
AtlasCloudatlas-cloud/fp8	$0.30	$1.20	$0.03	—
Baidubaidu/fp8	$0.30	$1.20	$0.006	—
BaseTenbaseten/fast	$0.60	$2.40	$0.014	—
BaseTenbaseten/fp8	$0.30	$1.20	$0.007	—
CoreWeavecoreweave/fp8	$0.20	$0.65	$0.03	—
Crusoecrusoe/fp8	$0.29	$1.20	$0.007	—
Decartdecart/fp4	$0.09	$0.18	$0.018	—
DeepInfradeepinfra/fp8	$0.14	$0.42	$0.0042	—
DeepSeekdeepseek	$0.15	$0.60	$0.003	—
DekaLLMdekallm	$0.12	$1.20	$0.005	—
DigitalOceandigitalocean	$0.165	$0.66	$0.006	—
Fireworksfireworks	$0.30	$1.20	$0.006	—
Fireworksfireworks/us	$0.45	$1.80	$0.009	—
GMICloudgmicloud/fp8	$0.18	$0.72	$0.0036	—
InferenceNetinference-net/fp8	$0.07	$0.60	$0.01	—
Ionstreamionstream	$0.10	$1.10	$0.01	—
Makoramakora/fp8	$0.27	$1.15	$0.009	—
Modalmodal	$0.30	$1.20	$0.03	—
Morphmorph/fp8	$0.029	$1.00	$0.009	—
Novitanovita/fp8	$0.195	$0.78	$0.0039	—
OpenInferenceopen-inference/fp4	$0.00011	$0.36	$0.00011	—
Parasailparasail/fp8	$0.30	$1.20	$0.006	—
Phalaphala	$0.21	$0.84	$0.0042	—
Relacerelace	$0.0001	$0.60	$0.005	—
Sail Researchsail-research/fp4	$0.08	$0.40	$0.01	—
SiliconFlowsiliconflow/fp8	$0.30	$1.20	$0.006	—
StreamLakestreamlake/fp8	$0.147	$0.588	$0.00294	—
Togethertogether	$0.30	$1.20	$0.006	—
Venicevenice/fp8	$0.30	$1.20	$0.0075	—
Waferwafer	$0.05	$1.60	$0.049	—

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

6,274–35,880.4 ms

AtlasCloudatlas-cloud/fp8

1,395.5–3,389.54 ms

Baidubaidu/fp8

797–7,358.96 ms

BaseTenbaseten/fast

381–7,974.58 ms

CoreWeavecoreweave/fp8

819–2,663.27 ms

Crusoecrusoe/fp8

616–5,475.88 ms

All latency percentiles 30 endpoints

Endpoint	P50	P75	P90	P99
Alibabaalibaba	6,274 ms	12,397.5 ms	18,910 ms	35,880.4 ms
AtlasCloudatlas-cloud/fp8	1,395.5 ms	1,702 ms	2,200 ms	3,389.54 ms
Baidubaidu/fp8	797 ms	1,101 ms	1,681.2 ms	7,358.96 ms
BaseTenbaseten/fast	381 ms	698.5 ms	1,292 ms	7,974.58 ms
CoreWeavecoreweave/fp8	819 ms	1,007 ms	1,331 ms	2,663.27 ms
Crusoecrusoe/fp8	616 ms	872.25 ms	1,268 ms	5,475.88 ms
Decartdecart/fp4	1,171 ms	6,278.25 ms	11,843.5 ms	20,772.86 ms
DeepInfradeepinfra/fp8	1,188 ms	1,785 ms	2,710.9 ms	7,648.18 ms
DeepSeekdeepseek	1,011 ms	1,324 ms	1,670 ms	2,915.36 ms
DekaLLMdekallm	838 ms	1,304 ms	1,973 ms	5,766.24 ms
DigitalOceandigitalocean	760.5 ms	1,088 ms	1,752.8 ms	4,659.27 ms
Fireworksfireworks	7,134.5 ms	21,270.75 ms	44,039.9 ms	89,550.65 ms
Fireworksfireworks/us	306 ms	589 ms	945.1 ms	2,464.93 ms
GMICloudgmicloud/fp8	2,535.5 ms	4,168.75 ms	8,551.5 ms	35,627.08 ms
InferenceNetinference-net/fp8	564 ms	973 ms	1,596 ms	4,661.54 ms
Ionstreamionstream	617 ms	871 ms	1,250 ms	3,745.2 ms
Makoramakora/fp8	684 ms	1,079.25 ms	1,614 ms	3,149.26 ms
Modalmodal	521 ms	759 ms	1,271 ms	5,099.42 ms
Morphmorph/fp8	1,002 ms	2,343 ms	5,022.9 ms	14,326.86 ms
Novitanovita/fp8	1,152 ms	1,799.25 ms	2,869.9 ms	4,697.36 ms
OpenInferenceopen-inference/fp4	2,502.5 ms	3,935 ms	8,333.2 ms	22,511.86 ms
Parasailparasail/fp8	895 ms	1,490.25 ms	2,209.9 ms	9,258.72 ms
Phalaphala	1,576 ms	2,125 ms	2,656.7 ms	4,440.95 ms
Relacerelace	976.5 ms	1,681.25 ms	2,791.9 ms	6,012.09 ms
Sail Researchsail-research/fp4	1,352 ms	1,656 ms	1,967 ms	3,242.45 ms
SiliconFlowsiliconflow/fp8	6,267.5 ms	13,435.5 ms	23,747.7 ms	62,869.43 ms
StreamLakestreamlake/fp8	1,759.5 ms	2,408 ms	3,445.8 ms	8,331.99 ms
Togethertogether	312 ms	465 ms	1,028 ms	9,443.73 ms
Venicevenice/fp8	970 ms	1,462 ms	2,049.9 ms	5,915.42 ms
Waferwafer	874 ms	1,251 ms	2,510.7 ms	5,444.43 ms

Output throughput · tokens per second

Throughput distributionRolling 30-minute snapshot · 6 endpoints

P50P75P90P99

Alibabaalibaba

39–132 t/s

AtlasCloudatlas-cloud/fp8

152–219.09 t/s

Baidubaidu/fp8

159–358.6 t/s

BaseTenbaseten/fast

292–782.64 t/s

CoreWeavecoreweave/fp8

224.5–367 t/s

Crusoecrusoe/fp8

135–240 t/s

All throughput percentiles 30 endpoints

Endpoint	P50	P75	P90	P99
Alibabaalibaba	39 t/s	66 t/s	87 t/s	132 t/s
AtlasCloudatlas-cloud/fp8	152 t/s	182 t/s	196 t/s	219.09 t/s
Baidubaidu/fp8	159 t/s	227 t/s	282 t/s	358.6 t/s
BaseTenbaseten/fast	292 t/s	429 t/s	598.4 t/s	782.64 t/s
CoreWeavecoreweave/fp8	224.5 t/s	266 t/s	302 t/s	367 t/s
Crusoecrusoe/fp8	135 t/s	165 t/s	191 t/s	240 t/s
Decartdecart/fp4	29 t/s	38 t/s	46 t/s	61 t/s
DeepInfradeepinfra/fp8	44 t/s	54 t/s	64 t/s	85 t/s
DeepSeekdeepseek	153 t/s	203 t/s	231 t/s	305 t/s
DekaLLMdekallm	148 t/s	202.75 t/s	246 t/s	317.17 t/s
DigitalOceandigitalocean	195 t/s	250 t/s	293 t/s	372 t/s
Fireworksfireworks	40 t/s	89 t/s	118 t/s	160 t/s
Fireworksfireworks/us	226 t/s	275 t/s	329 t/s	423.3 t/s
GMICloudgmicloud/fp8	103 t/s	148 t/s	171 t/s	212 t/s
InferenceNetinference-net/fp8	154 t/s	201 t/s	251 t/s	357 t/s
Ionstreamionstream	105 t/s	140 t/s	181 t/s	277.28 t/s
Makoramakora/fp8	142 t/s	190 t/s	230 t/s	304 t/s
Modalmodal	140 t/s	189 t/s	270 t/s	437 t/s
Morphmorph/fp8	85 t/s	127 t/s	173.9 t/s	254 t/s
Novitanovita/fp8	94 t/s	114 t/s	138 t/s	212 t/s
OpenInferenceopen-inference/fp4	19 t/s	22 t/s	24 t/s	33 t/s
Parasailparasail/fp8	181 t/s	270 t/s	345 t/s	484 t/s
Phalaphala	69 t/s	139 t/s	202 t/s	252 t/s
Relacerelace	85 t/s	110 t/s	132 t/s	180 t/s
Sail Researchsail-research/fp4	52 t/s	67 t/s	80 t/s	113.09 t/s
SiliconFlowsiliconflow/fp8	71 t/s	115 t/s	139 t/s	164 t/s
StreamLakestreamlake/fp8	97 t/s	154 t/s	176 t/s	214 t/s
Togethertogether	229 t/s	279 t/s	322 t/s	416 t/s
Venicevenice/fp8	143 t/s	255 t/s	328 t/s	447 t/s
Waferwafer	113 t/s	155 t/s	214 t/s	302.09 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 · 31 endpoints

Endpoint5 minutes30 minutes24 hours

AlibabaalibabaObserved Oct 10, 2026, 11:31 AM UTC

100%

95.56%

AtlasCloudatlas-cloud/fp8Observed Oct 10, 2026, 11:31 AM UTC

100%

Baidubaidu/fp8Observed Oct 10, 2026, 11:31 AM UTC

100%

98.99%

BaseTenbaseten/fastObserved Oct 10, 2026, 11:31 AM UTC

80%

96.67%

97.44%

BaseTenbaseten/fp8Observed Oct 10, 2026, 11:31 AM UTC

100%

CoreWeavecoreweave/fp8Observed Oct 10, 2026, 11:31 AM UTC

100%

99.72%

Crusoecrusoe/fp8Observed Oct 10, 2026, 11:31 AM UTC

100%

Decartdecart/fp4Observed Oct 10, 2026, 11:31 AM UTC

100%

96.67%

99.17%

DeepInfradeepinfra/fp8Observed Oct 10, 2026, 11:31 AM UTC

100%

DeepSeekdeepseekObserved Oct 10, 2026, 11:31 AM UTC

100%

DekaLLMdekallmObserved Oct 10, 2026, 11:31 AM UTC

100%

DigitalOceandigitaloceanObserved Oct 10, 2026, 11:31 AM UTC

100%

99.58%

FireworksfireworksObserved Oct 10, 2026, 11:31 AM UTC

20%

3.33%

55.35%

Fireworksfireworks/usObserved Oct 10, 2026, 11:31 AM UTC

100%

GMICloudgmicloud/fp8Observed Oct 10, 2026, 11:31 AM UTC

100%

99.93%

InferenceNetinference-net/fp8Observed Oct 10, 2026, 11:31 AM UTC

100%

IonstreamionstreamObserved Oct 10, 2026, 11:31 AM UTC

100%

99.93%

Makoramakora/fp8Observed Oct 10, 2026, 11:31 AM UTC

100%

ModalmodalObserved Oct 10, 2026, 11:31 AM UTC

100%

Morphmorph/fp8Observed Oct 10, 2026, 11:31 AM UTC

100%

Novitanovita/fp8Observed Oct 10, 2026, 11:31 AM UTC

100%

OpenInferenceopen-inference/fp4Observed Oct 10, 2026, 11:31 AM UTC

60%

73.33%

88.2%

Parasailparasail/fp8Observed Oct 10, 2026, 11:31 AM UTC

100%

PhalaphalaObserved Oct 10, 2026, 11:31 AM UTC

100%

RelacerelaceObserved Oct 10, 2026, 11:31 AM UTC

100%

Sail Researchsail-research/fp4Observed Oct 10, 2026, 11:31 AM UTC

100%

SiliconFlowsiliconflow/fp8Observed Oct 10, 2026, 11:31 AM UTC

100%

99.79%

StreamLakestreamlake/fp8Observed Oct 10, 2026, 11:31 AM UTC

100%

TogethertogetherObserved Oct 10, 2026, 11:31 AM UTC

100%

99.58%

Venicevenice/fp8Observed Oct 10, 2026, 11:31 AM UTC

100%

WaferwaferObserved Oct 10, 2026, 11:31 AM UTC

100%

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-v4.1-flashObserved Oct 10, 2026, 11:31 AM UTC

Benchmark profileArtificial Analysis · shared source scale from 0 to 40

Intelligence

39.5

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 V4.1 Flash

87.879%

gpqa diamondDeepSeek: DeepSeek V4.1 Flash

89.091%

tau bench verified airlineDeepSeek: DeepSeek V4.1 Flash

76.8%

Evaluation	Configuration	Source	Results	Details
gpqa diamond	Deepseek V4.1 Flashdeepseek/deepseek-v4.1-flash	OpenRouter	87.879%	Details
gpqa diamond	DeepSeek: DeepSeek V4.1 Flashdeepseek/deepseek-v4.1-flash-20260910	OpenRouter	89.091%	Details
Composite indices	DeepSeek V4.1 Flash (Non-reasoning)deepseek/deepseek-v4.1-flash-20260910	Artificial Analysis	Intelligence 24.7 · Coding — · Agentic —	Details
tau bench verified airline	DeepSeek: DeepSeek V4.1 Flashdeepseek/deepseek-v4.1-flash-20260910	OpenRouter	76.8%	Details

Sources: Artificial Analysis, Design Arena and OpenRouter. Evaluation configurations and scales differ; results are not interchangeable. Updated Oct 24, 2026.

Apps & session costs

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

30-day sample · through Oct 18, 2026

Application	Model route	Turns	Median cost / session	Details
Kilo Code	deepseek/deepseek-v4.1-flash-20260910	50+ turns	$0.27109	Details
Claude Code	deepseek/deepseek-v4.1-flash-20260910	2–9 turns	$0.008579	Details
Claude Code	deepseek/deepseek-v4.1-flash-20260910	10–49 turns	$0.047038	Details
Claude Code	deepseek/deepseek-v4.1-flash-20260910	1 turn	$0.003223	Details
Codex	deepseek/deepseek-v4.1-flash-20260910	50+ turns	$0.10392	Details
Kilo Code	deepseek/deepseek-v4.1-flash-20260910	2–9 turns	$0.011654	Details
Claude Code	deepseek/deepseek-v4.1-flash-20260910	50+ turns	$0.343792	Details
Kilo Code	deepseek/deepseek-v4.1-flash-20260910	1 turn	$0.001209	Details
Codex	deepseek/deepseek-v4.1-flash-20260910	2–9 turns	$0.005091	Details
Hermes Agent	deepseek/deepseek-v4.1-flash-20260910	1 turn	$0.00082	Details
Codex	deepseek/deepseek-v4.1-flash-20260910	1 turn	$0.000735	Details
Codex	deepseek/deepseek-v4.1-flash-20260910	10–49 turns	$0.031868	Details
Hermes Agent	deepseek/deepseek-v4.1-flash-20260910	10–49 turns	$0.045058	Details
Hermes Agent	deepseek/deepseek-v4.1-flash-20260910	50+ turns	$0.271857	Details
Hermes Agent	deepseek/deepseek-v4.1-flash-20260910	2–9 turns	$0.005639	Details
Kilo Code	deepseek/deepseek-v4.1-flash-20260910	10–49 turns	$0.045824	Details

Source: OpenRouter, as of Oct 20, 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 usage153,532,244,497,479 reported tokens · 44 days with observations

7.5T0

2026-09-10
678,686,144,342 tokens

2026-09-11
1,536,367,792,104 tokens

2026-09-12
1,428,066,412,438 tokens

2026-09-13
1,294,618,265,169 tokens

2026-09-14
1,283,244,821,904 tokens

2026-09-15
1,799,484,922,582 tokens

2026-09-16
2,249,099,456,858 tokens

2026-09-17
2,250,990,777,624 tokens

2026-09-18
2,621,995,119,219 tokens

2026-09-19
3,072,929,126,131 tokens

2026-09-20
2,490,889,275,600 tokens

2026-09-21
2,431,719,459,493 tokens

2026-09-22
2,717,219,458,442 tokens

2026-09-23
2,809,507,461,498 tokens

2026-09-24
2,761,232,489,714 tokens

2026-09-25
2,964,159,850,392 tokens

2026-09-26
2,451,574,001,771 tokens

2026-09-27
3,440,258,106,142 tokens

2026-09-28
3,681,553,673,810 tokens

2026-09-29
3,918,021,317,885 tokens

2026-09-30
3,491,740,784,985 tokens

2026-10-01
3,549,428,106,767 tokens

2026-10-02
3,641,016,456,262 tokens

2026-10-03
3,908,062,559,099 tokens

2026-10-04
3,419,621,759,703 tokens

2026-10-05
6,458,330,218,100 tokens

2026-10-06
5,107,537,657,195 tokens

2026-10-07
7,519,183,440,956 tokens

2026-10-08
5,537,233,011,384 tokens

2026-10-09
5,648,313,383,757 tokens

2026-10-10
5,022,780,525,937 tokens

2026-10-11
5,679,385,293,681 tokens

2026-10-12
6,020,046,730,509 tokens

2026-10-13
5,046,621,695,407 tokens

2026-10-14
4,474,714,742,293 tokens

2026-10-15
5,167,300,442,116 tokens

2026-10-16
5,413,760,580,051 tokens

2026-10-17
4,086,846,875,053 tokens

2026-10-18
2,623,123,996,518 tokens

2026-10-19
2,803,399,789,975 tokens

2026-10-20
2,761,395,726,132 tokens

2026-10-21
2,856,246,398,895 tokens

2026-10-22
2,930,966,988,493 tokens

2026-10-23
2,483,569,401,093 tokens

Jul 26, 2026Oct 23, 2026

Daily observations · 44 days

Date (UTC)	Reported tokens	Reported routes
Oct 23, 2026	2,483,569,401,093	1
Oct 22, 2026	2,930,966,988,493	1
Oct 21, 2026	2,856,246,398,895	1
Oct 20, 2026	2,761,395,726,132	1
Oct 19, 2026	2,803,399,789,975	1
Oct 18, 2026	2,623,123,996,518	1
Oct 17, 2026	4,086,846,875,053	1
Oct 16, 2026	5,413,760,580,051	1
Oct 15, 2026	5,167,300,442,116	1
Oct 14, 2026	4,474,714,742,293	1
Oct 13, 2026	5,046,621,695,407	1
Oct 12, 2026	6,020,046,730,509	1
Oct 11, 2026	5,679,385,293,681	1
Oct 10, 2026	5,022,780,525,937	1
Oct 9, 2026	5,648,313,383,757	1
Oct 8, 2026	5,537,233,011,384	1
Oct 7, 2026	7,519,183,440,956	1
Oct 6, 2026	5,107,537,657,195	1
Oct 5, 2026	6,458,330,218,100	1
Oct 4, 2026	3,419,621,759,703	1
Oct 3, 2026	3,908,062,559,099	1
Oct 2, 2026	3,641,016,456,262	1
Oct 1, 2026	3,549,428,106,767	1
Sep 30, 2026	3,491,740,784,985	1
Sep 29, 2026	3,918,021,317,885	1
Sep 28, 2026	3,681,553,673,810	1
Sep 27, 2026	3,440,258,106,142	1
Sep 26, 2026	2,451,574,001,771	1
Sep 25, 2026	2,964,159,850,392	1
Sep 24, 2026	2,761,232,489,714	1
Sep 23, 2026	2,809,507,461,498	1
Sep 22, 2026	2,717,219,458,442	1
Sep 21, 2026	2,431,719,459,493	1
Sep 20, 2026	2,490,889,275,600	1
Sep 19, 2026	3,072,929,126,131	1
Sep 18, 2026	2,621,995,119,219	1
Sep 17, 2026	2,250,990,777,624	1
Sep 16, 2026	2,249,099,456,858	1
Sep 15, 2026	1,799,484,922,582	1
Sep 14, 2026	1,283,244,821,904	1
Sep 13, 2026	1,294,618,265,169	1
Sep 12, 2026	1,428,066,412,438	1
Sep 11, 2026	1,536,367,792,104	1
Sep 10, 2026	678,686,144,342	1

Source: OpenRouter (openrouter.ai/rankings), as of Oct 24, 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-v4.1-flashObserved Oct 10, 2026, 11:31 AM UTC

Context: 1,048,576
Maximum output: 943,718
Tokenizer: DeepSeek
Input: text, image
Output: text
Moderation: No
Added to OpenRouter: 2026-09-10

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

Model IDdeepseek/deepseek-v4.1-flash

Version IDdeepseek/deepseek-v4.1-flash-20260910

Weights IDdeepseek-ai/DeepSeek-V4.1-Flash

Reasoning

Required: No
Enabled by default: Yes
Default effort: high
Supported efforts: max, high, low

Supported parameters

frequency_penalty
include_reasoning
logit_bias
logprobs
max_tokens
min_p
presence_penalty
reasoning
reasoning_effort
repetition_penalty
response_format
seed
stop
structured_outputs
temperature
tool_choice
tools
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 V4.1 Flash cost?

Listed rates start at $0.0001 per million input tokens and $0.18 per million output tokens. Prices vary by provider and configuration; see Pricing for details.

Which providers offer DeepSeek V4.1 Flash?

There are 129 cataloged offers across 68 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: Not reported. Structured output: Not reported. Support can vary by endpoint.

Resources

MIT

Hugging Face ↗

Catalog history · 4 recorded revisions

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

Oct 20, 2026 · 7b89b3be3684
Sep 19, 2026 · 7b88df1d426f
Sep 14, 2026 · b48f8c44b42b
Sep 12, 2026 · a979c5c2e8fd