DeepSeek V4.1 Flash

New

DeepSeekReleased Sep 10, 2026

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.

Try it in the estimate Compare

Input, per 1M tokens: $0.36$0.33 for holders
Output, per 1M tokens: $1.44$1.32 for holders
Context: 1M tokens
Tokens this week: None yet

Price

Agents pay these prices from their AI budget. $BINF holders pay up to 8% less for the credit they buy.

Charge	Agents	Holders	Per
Inputper 1M tokens	$0.36	$0.33	per 1M tokens
Outputper 1M tokens	$1.44	$1.32	per 1M tokens
Cached inputper 1M tokens	$0.0072	$0.0066	per 1M tokens

Price history

What an agent paid per million tokens, by day.

Unchanged since Sep 29, 2026: $0.36 in and $1.44 out per million tokens.

Usage through bInference

Every call our agents make to this model, by day. Counts only: no prompt or answer is stored.

Last 7 days: 0
Last 30 days: 0
All time: 0

InputOutputReasoning

Where it runs

The hosts serving this model now, and how often each answered over the last day.

Use it

Two settings: our address and your agent’s key. The model is set in each call.

import OpenAI from "openai";const client = new OpenAI({  baseURL: "https://binference.io/api/v1",Address  apiKey: process.env.BINF_API_KEY,Key});const reply = await client.chat.completions.create({  model: "deepseek/deepseek-v4.1-flash",  messages: [{ role: "user", content: "Summarize today's trades" }],  max_tokens: 1024,});

curl https://binference.io/api/v1/chat/completions \Address  -H "Authorization: Bearer $BINF_API_KEY" \Key  -H "Content-Type: application/json" \  -d '{    "model": "deepseek/deepseek-v4.1-flash",    "messages": [{ "role": "user", "content": "Hello" }],    "max_tokens": 1024  }'

Questions

More from DeepSeek

DeepSeek V4 Flash Vision ExpDeepSeek1M$0.26 in · $0.78 out$0.24 in · $0.71 out DeepSeek V4 Pro 0813DeepSeek1M$0.47 in · $4.19 out$0.44 in · $3.85 out DeepSeek V4 Flash 0731DeepSeek1M$0.022 in · $0.38 out$0.020 in · $0.35 out DeepSeek V4 Pro 0423DeepSeek1M$1.10 in · $2.20 out$1.01 in · $2.02 out

Agents

Holders

Inputper 1M tokens

$0.36

$0.33

per 1M tokens

Outputper 1M tokens

$1.44

$1.32

per 1M tokens

Cached inputper 1M tokens

$0.0072

$0.0066

per 1M tokens

Host	Answered, last day	Context
Relace, answering	99.9%	1M
OpenInferenceFp4fp4, answering	99.1%	1M
InferenceNet, answering	99%	1M
Sail ResearchFp4fp4, having trouble	95.7%	1M
MorphFp8fp8, answering	96.5%	1M
Wafer, answering	99.9%	1M
AtlasCloudFp8fp8, answering	99.8%	1M
DekaLLM, answering	97.2%	1M
DeepInfraFp8fp8, answering	99.9%	1M
Ionstream, answering	96.3%	1M
Alibaba, answering	99.7%	1M
DeepSeek, answering	100%	1M
StreamLakeFp8fp8, answering	99.6%	1M
GMICloudFp8fp8, answering	99.7%	1M
NovitaFp8fp8, answering	99.9%	1M
CoreWeaveFp8fp8, answering	98.9%	1M
NextBitFp8fp8, answering	99.8%	1M
Phala, answering	99.7%	1M
Fireworks, answering	99%	1M
KreaFp8fp8, answering	99%	1M
Io NetFp8fp8, answering	99.8%	1M
BaiduFp8fp8, answering	99.4%	1M
BaseTenFp8fp8, answering	99.2%	1M
MakoraFp8fp8, answering	99.9%	1M
DigitalOcean, answering	99.6%	1M
Together, answering	99.9%	1M
SiliconFlowFp8fp8, answering	99.8%	1M
Modal, answering	99.8%	1M
BaseTenFp8fp8, answering	98.7%	1M
ParasailFp8fp8, answering	99.4%	1M
VeniceFp8fp8, answering	94.7%	1M
FireworksUS, answering	99.8%	1M
BaseTenFastfp32, answering	99.9%	1M

Host

Answered, last day

Context

Relace

99.9%

OpenInferenceFp4fp4

99.1%

InferenceNet

99%

Sail ResearchFp4fp4

95.7%

MorphFp8fp8

96.5%

Wafer

99.9%

AtlasCloudFp8fp8

99.8%

DekaLLM

97.2%

DeepInfraFp8fp8

99.9%

Ionstream

96.3%

Alibaba

99.7%

DeepSeek

100%

StreamLakeFp8fp8

99.6%

GMICloudFp8fp8

99.7%

NovitaFp8fp8

99.9%

CoreWeaveFp8fp8

98.9%

NextBitFp8fp8

99.8%

Phala

99.7%

Fireworks

99%

KreaFp8fp8

99%

Io NetFp8fp8

99.8%

BaiduFp8fp8

99.4%

BaseTenFp8fp8

99.2%

MakoraFp8fp8

99.9%

DigitalOcean

99.6%

Together

99.9%

SiliconFlowFp8fp8

99.8%

Modal

99.8%

BaseTenFp8fp8

98.7%

ParasailFp8fp8

99.4%

VeniceFp8fp8

94.7%

FireworksUS

99.8%

BaseTenFastfp32

99.9%

import OpenAI from "openai";const client = new OpenAI({ baseURL: "https://binference.io/api/v1",Address apiKey: process.env.BINF_API_KEY,Key});const reply = await client.chat.completions.create({ model: "deepseek/deepseek-v4.1-flash", messages: [{ role: "user", content: "Summarize today's trades" }], max_tokens: 1024,});

curl https://binference.io/api/v1/chat/completions \Address -H "Authorization: Bearer $BINF_API_KEY" \Key -H "Content-Type: application/json" \ -d '{ "model": "deepseek/deepseek-v4.1-flash", "messages": [{ "role": "user", "content": "Hello" }], "max_tokens": 1024 }'

DeepSeek V4.1 Flash

Price

Price history

Usage through bInference

Where it runs

Use it

Questions

More from DeepSeek

Similar models

DeepSeek V4.1 Flash

Price

Price history

Usage through bInference

Where it runs

Where it runs

Use it

Questions

More from DeepSeek

Similar models

Where it runs

DeepSeek V4.1 Flash

Price

Price history

Usage through bInference

Where it runs

Use it

Questions

How much does DeepSeek V4.1 Flash cost?

How does my agent use DeepSeek V4.1 Flash?

How much can DeepSeek V4.1 Flash read at once?

What can DeepSeek V4.1 Flash do?

Are my prompts stored?

More from DeepSeek

Similar models

DeepSeek V4.1 Flash

Price

Price history

Usage through bInference

Where it runs

Where it runs

Use it

Questions

How much does DeepSeek V4.1 Flash cost?

How does my agent use DeepSeek V4.1 Flash?

How much can DeepSeek V4.1 Flash read at once?

What can DeepSeek V4.1 Flash do?

Are my prompts stored?

More from DeepSeek

Similar models

Where it runs