About DeepSeek V4.1 Flash

DeepSeek-V4.1-Flash est le modèle phare allégé de la nouvelle famille d’architectures DeepSeek. Doté d’un total de 552 milliards de paramètres MoE, il offre des performances intellectuelles surpassant celles des modèles phares dans les principaux benchmarks, notamment en surpassant DeepSeek-V4-Pro. Il adopte une conception asymétrique de type « Causal Encoder-Decoder » avec 8 milliards de paramètres actifs en entrée et 16 milliards en sortie, et offre une compréhension visuelle multimodale native. L’utilisation du cache KV est réduite à un quart de la mémoire HBM de la génération précédente et à un huitième de son stockage SSD, ce qui diminue considérablement les coûts pour les charges de travail à contexte long et les tâches agentiques. Avec une fenêtre de contexte de 1 million de tokens et une sortie maximale de 384 000 tokens, il offre un équilibre entre un débit élevé, une faible latence et une rentabilité exceptionnelle.

DeepSeek V4.1 Flash est un modèle conversationnel développé par DeepSeek, disponible sur Eden AI via 18 endpoints chez 14 fournisseurs, à partir de 0,098 $ par million de tokens en entrée et 0,39 $ par million de tokens en sortie, avec une fenêtre de contexte allant jusqu'à 1,05M tokens.

Providers and prices

18 endpoints from 14 providers, cheapest first, in USD per million tokens. Every one is called with the same Eden AI key: pin one with its endpoint ID, or let Eden AI route for you. .

Endpoint

Region

Context

Input / 1M

Output / 1M

Cached / 1M

Supports

QwenOfficial−35%qwen/deepseek-v4.1-flash

Monde

1M

$0.15$0.0975

$0.39

$0.0098

ReasoningToolsJSONCacheVisionWeb

FlexAIOfficial−%flexai/DeepSeek-V4.1-Flash

Monde

1.05M

$0.1$0.1

$0.66

$0.015

ReasoningToolsJSONCacheVisionWeb

DeepSeek (Asie)Official−%deepseek/deepseek-flash

Asie

1M

$0.15$0.15

$0.6

$0.003

ReasoningToolsJSONCacheVisionWeb

DeepSeek (Asie)Official−%deepseek/deepseek-v4-flash

Asie

1M

$0.15$0.15

$0.6

$0.003

ReasoningToolsJSONCacheVisionWeb

DeepSeek (Asie)Official−%deepseek/deepseek-v4-flash-vision-exp

Asie

1M

$0.15$0.15

$0.6

$0.003

ReasoningToolsJSONCacheVisionWeb

Tencent Cloud (Asie)Official−%tencent/deepseek-v4.1-flash

Asie

1M

$0.15$0.15

$0.6

$0.003

ReasoningToolsJSONCacheVisionWeb

DeepInfra (États-Unis)Official−%deepinfra/deepseek-ai/DeepSeek-V4.1-Flash

États-Unis

1.05M

$0.2$0.2

$0.6

$0.006

ReasoningToolsJSONCacheVisionWeb

Greenference (Europe)Official−%greenference/deepseek-v4.1-flash

Europe

1.05M

$0.22$0.22

$0.88

$0.007

ReasoningToolsJSONCacheVisionWeb

GreenPT (Europe)Official−%greenpt/deepseek-v4.1-flash

Europe

1M

$0.2476$0.2476

$1.2378

$0.0124

ReasoningToolsJSONCacheVisionWeb

DatabricksOfficial−%databricks/databricks-deepseek-v4-1-flash

Monde

$0.3$0.3

$1.2

$0.03

ReasoningToolsJSONCacheVisionWeb

NebiusOfficial−%nebius/deepseek-ai/DeepSeek-V4.1-Flash

Monde

1.05M

$0.3$0.3

$1.2

$0.3

ReasoningToolsJSONCacheVisionWeb

Fireworks AI (États-Unis)Official−%fireworks_ai/accounts/fireworks/models/deepseek-v4p1-flash

États-Unis

1.05M

$0.3$0.3

$1.2

$0.006

ReasoningToolsJSONCacheVisionWeb

Together AIOfficial−%together_ai/deepseek-ai/DeepSeek-V4.1-Flash

Monde

1.05M

$0.3$0.3

$1.2

$0.006

ReasoningToolsJSONCacheVisionWeb

Google CloudOfficial−%vertex/deepseek-ai/deepseek-v4.1-flash-maas

Monde

$0.3$0.3

$1.2

$0.03

ReasoningToolsJSONCacheVisionWeb

Google Cloud (Europe)Official−%vertex/deepseek-ai/deepseek-v4.1-flash-maas@eu

Europe

$0.33$0.33

$1.32

$0.033

ReasoningToolsJSONCacheVisionWeb

Google Cloud (États-Unis)Official−%vertex/deepseek-ai/deepseek-v4.1-flash-maas@us

États-Unis

$0.33$0.33

$1.32

$0.033

ReasoningToolsJSONCacheVisionWeb

TensorX (Europe)Official−%tensorx/deepseek/deepseek-v4.1-flash

Europe

1.05M

$0.5$0.5

$1.5

$0.125

ReasoningToolsJSONCacheVisionWeb

Scaleway (Europe)Official−%scaleway/deepseek-v4.1-flash

Europe

256K

$0.6752$0.6752

$2.0255

$0.135

ReasoningToolsJSONCacheVisionWeb

Performance

Measured on real Eden AI traffic over the last 30 days. Endpoints with too little traffic are not shown.

Fastest first token

0.46 s

Together AI

Lowest latency

0.55 s

TensorX (Europe)

Highest throughput

197.36 tok/s

Together AI

Best availability

100%

Scaleway (Europe)

Endpoint

First token

Latency p50

Latency p95

Throughput

Availability

Cache hits

Tool errors

JSON errors

Qwen

3.07 s

8.45 s

38.82 s

79.54 tok/s

99.95%

87.33%

2.85%

70.39%

FlexAI

3.15 s

3.85 s

14.49 s

75.46 tok/s

99.64%

94.4%

0.57%

0.23%

DeepSeek (Asie)

0.62 s

3.61 s

17.48 s

144.57 tok/s

99.91%

87.07%

0.44%

88.04%

DeepSeek (Asie)

0.71 s

2.22 s

13.09 s

93.35 tok/s

99.88%

97.51%

0.16%

5.64%

DeepSeek (Asie)

s

tok/s

%

Tencent Cloud (Asie)

1.96 s

3.31 s

66.53 s

89.29 tok/s

100%

92.82%

0%

DeepInfra (États-Unis)

1.46 s

4.06 s

21.56 s

49.34 tok/s

99.57%

92.54%

0.44%

0.64%

Greenference (Europe)

0.99 s

4.01 s

49.55 s

131.55 tok/s

99.82%

97.48%

12.95%

70.97%

GreenPT (Europe)

0.9 s

2.71 s

54.05 s

82.93 tok/s

95.2%

91.37%

8.35%

78.57%

Databricks

1.94 s

6.18 s

28.56 s

121.57 tok/s

99.48%

9.36%

1.35%

2.44%

Nebius

2.61 s

4.17 s

30.49 s

110.18 tok/s

99.91%

89.73%

0%

1.85%

Fireworks AI (États-Unis)

1.39 s

3.68 s

31.77 s

64.18 tok/s

99.96%

96.89%

0.06%

7.69%

Together AI

0.46 s

1.89 s

16 s

197.36 tok/s

97.31%

90.76%

9.01%

2.94%

Google Cloud

1.73 s

7.19 s

166.85 s

37.34 tok/s

93.94%

0%

Google Cloud (Europe)

1.73 s

6.61 s

89.12 s

46.1 tok/s

93.1%

0%

Google Cloud (États-Unis)

s

tok/s

%

TensorX (Europe)

1.81 s

0.55 s

4.37 s

2.09 tok/s

99.97%

82.31%

0.55%

0%

Scaleway (Europe)

1.54 s

0.92 s

29.4 s

118.71 tok/s

100%

53.87%

10.71%

0%

Estimate your monthly cost

Enter your expected volume to compare every endpoint at once.

Call DeepSeek V4.1 Flash with Eden AI

Use the model ID to let Eden AI pick an endpoint, or an endpoint ID from the table above to pin one.

import requests

response = requests.post(
    "https://api.edenai.run/v3/chat/completions",
    headers={"Authorization": "Bearer YOUR_EDENAI_API_KEY"},
    json={
        "model": "",
        "messages": [{"role": "user", "content": "Hello!"}],
    },
)
print(response.json())

curl https://api.edenai.run/v3/chat/completions \
  -H "Authorization: Bearer YOUR_EDENAI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model": "", "messages": [{"role": "user", "content": "Hello!"}]}'

const response = await fetch("https://api.edenai.run/v3/chat/completions", {
  method: "POST",
  headers: {
    Authorization: "Bearer YOUR_EDENAI_API_KEY",
    "Content-Type": "application/json",
  },
  body: JSON.stringify({
    model: "",
    messages: [{ role: "user", content: "Hello!" }],
  }),
});
console.log(await response.json());

Questions about DeepSeek V4.1 Flash

How much does DeepSeek V4.1 Flash cost?

On Eden AI, prices for this model start at $0.0975 per million input tokens and $0.39 per million output tokens. Prices differ by endpoint; the providers table lists each one. .

What is the context window of DeepSeek V4.1 Flash?

Up to 1.05M tokens, depending on the endpoint.

Which providers serve DeepSeek V4.1 Flash?

18 endpoints from 14 providers, available in: Monde, Asie, Europe, États-Unis. All of them are called with the same Eden AI API key.

Can I run DeepSeek V4.1 Flash in the EU?

Yes. At least one endpoint is hosted in Europe; pin it with its endpoint ID from the providers table to keep requests in the EU.

How fast is DeepSeek V4.1 Flash?

Over the last 30 days, the fastest first token was 0.46 s (Together AI) and the highest throughput was 197.36 tokens per second (Together AI).

More from DeepSeek

Other DeepSeek models you can call with the same key.

See all DeepSeek models

Aucun élément.