DeepSeek V4 Flash

DeepSeek V4 Flash is DeepSeek's efficiency-optimized mixture-of-experts model, designed for fast inference on coding, reasoning, and agent workloads.

start building →

compare models

                                   . ......           .:-.
                        ... .   ..::-+**=**:.        .-%%*:.
                ..:---+++++--:--+==%@@@@@%*-.       .:=%@%+...                .
           ...-+==%@@@@@#@@@@@@@@##@@##@%+:.        .:%@@@%*-:..            .-+:
          .-*%@#@#@#@#@########@#@@@#@@@%-.         .-%@@@%%@%*-..     ....-*%=+.
       ..:=@##@####@####@@@@##@@@@@##@@@@=-:.       .-%@@@@%@%%%+:..-+****=%%%%+..
      .:+%###@##@@#@#@@##@@@#@@##@#@@#@@@@%*-:       :=%@@@@@%%%%**=%%%%%%%%%%=-.
     .-=@##@@#####@##@###@#@@@#@#@@@@@@@@@@@@=-:.    :+%@@@%@@@%%%%@@%%%%%%%%%*:
   ..-=###@#@#@#####@@#@####@@#@@@@@@@@@@@@@@@@*-..  .:*%%@@@@%@@%@%@%%%@%%%%*:..
   .-%@#@@@####@@##@@###@##@#@@@@#@@@@@@@@@@@@@@%*:.  .:+%%@%%%%@@@@@@%%%%%=*:..
  .:=@#@#@#@@######@@#@##@@@##@@@@@@@@@@@@@@@@@@@%=+:.  :-*%%@%%%%%@%%%%%==-.
  :+@@#@###@#@@@@@@##@@#@#@##@@@@@@@@@@@@@@@@@@@@@@%*-.. ..+%%%@@%%%%%=**-.
 .-%@@##@@@@%@@@@@#@@#@#@@@@@@@@@@@@@@@@@@@@@@@@%%%@%%*::.:-%@%@%%=*--:..
..+@#@@%+-:::::-+**=%@@@@@#@@@@@@@@@@@@@@@@@%%%@@@@%%%%%=*+=%@%%%%=+..
 :*@@#@=- .    .. .::-*%@@@@@@@@@@@@@@@@@@%*+-*=%@@@@%%%%%%%%%%%%%=-.
 :=@#@@%-.           ...:*%@@@@@@@@@@@@@@@%*+::.-*%%%@@@%%%%@%%%%%=:
.:=@@@@%-.               .:-=@@@@@@@@@@@@%%==%+...-%%%@%@%%%%%%%%%*:
 :=@##@@*.                 .:-=@@@@@@@@@@@@*=%*:. .-=%%@@@%%%%%%%=:.
..*@#@@@%:.                  ..+%@@@@@@@%%@%%%=:.  .-=%%%%%%%%%%=*..
 .+@##@@@+:                    .:*%@@@@@@@%@%%%*:...:+%%%%%%%%%%=-..
 .:%@#@@@%:.                    ..+%@@@%%%@@%%%%%****%%%%%%%%%%=*.
  .*@@@#@@=:.                     .-%%%%%%%@%@%%%%%%%%%%%%%%%%%=:.
  .-%@@@#@@*:.                     .:=%@@%@%%@%%%%%%%%%%%%%%%==-.
  ..+@@@@@@@*:.                    ..:*%%@%%@%%%%@%%%%%%%%%%==-.
   .:*%@@#@@@=:.           ... .     ..*%%%%%%@%%%%%%%%%%%==*-.
    .:*@@@@@@@=-:        ..-*+-...     :+=%%%%%%%%%%%%%%%%=*:.
    ..:*@@@@@@@%*:..     ..+%%%=*-:     :-=%%%%%%%%%%%%===+:.
       :+%@@@@@@%%+:.     .-=%%%%%*+:.  ..:*%%%%%%%%%%%==-..
        .-=@@@@@@%@%*-.....:+%%%@%%%=+:..  .-=%%%%%%%%%==++-....
         .:+%%@@@@@@%%=*+--+=%%%%%%%%%%+::...:+*%%=%==%=%===*+-.
           .:*%%%@@@@@%@%%%%%%%%@%%%%%%%%*++----*=%%=%%=%====**-.
             .:+=%%%@%@@@@%@%%@%%@%%%%%%%%%%====*+--+****=***+-.
               .:-*=%%%%%@%@@%@%%%%%%%%%%%%%==*+:............. .
                  .:-+*=%%%%%%%%%%%%%%%==%=*+:.
                    ....:-++***=******+--:..
                             .........

input · usd / 1m tok

$0.19

output · usd / 1m tok

$0.51

context window

1mtokens

modalities · in → out

pricing

base api ratesusd per 1m tokens

input: $0.19
output: $0.51
cache read: $0.028

providers

Azure AI Foundry hosts DeepSeek V4 Flash. Impossibl routes your request intelligently and handles failover automatically when a provider is down.

Azure AI Foundry
Azure AI Foundry serving options
endpoint location routes
East US 2 1
East US 1
locations describe provider endpoints; they do not guarantee where inference runs.

Azure AI Foundry serving options
endpoint location	routes
East US 2	1
East US	1

specifications

takes text in, gives text outdeepseek/deepseek-v4-flash

creatorDeepSeek
model typeText generation
context window1,000,000 tokens
maximum outputnot published
input
takes text in, gives text out
outputtext
tool callingnot verified
releasedapr 24, 2026
endpoint/v1/chat/completions

unknown means the public catalog does not establish that fact. a model's documented capabilities can differ from what a particular api supports. see the model documentation for integration details.

try it · your prompt, this model, live

# responses API
import os
from openai import OpenAI

client = OpenAI(
    api_key=os.environ["IMPOSSIBL_API_KEY"],
    base_url="https://api.impossibl.com/v1",
)

resp = client.responses.create(
    model="deepseek/deepseek-v4-flash",
    input="▌",
)
print(resp.output_text)

edit the prompt, then run

# responses API
import os
from openai import OpenAI

client = OpenAI(
    api_key=os.environ["IMPOSSIBL_API_KEY"],
    base_url="https://api.impossibl.com/v1",
)

resp = client.responses.create(
    model="deepseek/deepseek-v4-flash",
    input="▌",
)
print(resp.output_text)

edit the prompt, then run

benchmarks

intelligence index

34.3#39 of 108 models on the gateway

coding index69.1#25 of 64
agentic index41.0#21 of 53
omniscience-14.3#55 of 84
gpqa90.8%#22 of 80
humanity’s last exam38.6%#28 of 87

GPQA90.8%
AA-LCR79.7%
AutomationBench Partial54.0%
SciCode50.3%
Tau Banking39.4%
Humanity’s Last Exam38.6%
CritPt16.6%
Multilingual LCR13.3%
GDP.pdf All-pass11.0%

Intelligence Index34.33
Omniscience Index-14.28
GDPval Elo1433.97
legal index (API)37.7
coding index (API)69.1
agentic index (API)41
economics index (API)40.3
engineering index (API)34.9
intelligence index (API)34.3
strategy and ops index (API)43.1
finance and accounting index (API)37.5
healthcare and medical index (API)35.4

view source ↗all source data and breakdowns (json) ↗

each configuration is listed separately; these evaluations describe the tested model configuration, and performance and costs can differ on impossibl. unmeasured scores are omitted and api values keep their source units. last checked 2026-10-14.

compare DeepSeek V4 Flash with other models

frontier labs and open weights, best-ranked firstusd / 1m tok · in / out

anthropic/claude-opus-5-5$4.00 / $20.00compare →
openai/gpt-6-astra$10.00 / $50.00compare →
meta/muse-wonder$4.00 / $20.00compare →
xiaomi/mimo-v2.6-pro$0.435 / $0.87compare →

impossibl models --all →

frequently asked questions

What is DeepSeek V4 Flash?

DeepSeek V4 Flash is DeepSeek's efficiency-optimized mixture-of-experts model, designed for fast inference on coding, reasoning, and agent workloads. It pairs a large total parameter count with a small active set, supports long inputs, and later gained an experimental vision variant before being superseded by V4.1 Flash.

How much does DeepSeek V4 Flash cost?

DeepSeek V4 Flash costs $0.19 per million input tokens and $0.51 per million output tokens through the Impossibl API. Cache reads cost $0.028 per million tokens. All rates are in USD.

view the pricing table ↑

What is the context length of DeepSeek V4 Flash?

DeepSeek V4 Flash has a 1,000,000-token context window through the Impossibl API.

What inputs and outputs does DeepSeek V4 Flash support?

DeepSeek V4 Flash accepts text as input. DeepSeek V4 Flash returns text.

How do I use DeepSeek V4 Flash through an API?

DeepSeek V4 Flash is available through the Impossibl API. Use deepseek/deepseek-v4-flash as the model identifier with /v1/chat/completions. The quickstart explains how to create an API key and send your first request.

follow the api quickstart →

When was DeepSeek V4 Flash released?

DeepSeek released DeepSeek V4 Flash on April 24, 2026.

Creator release announcement ↗

this page describes DeepSeek V4 Flash as listed by the impossibl public api. prices and availability come from that catalog, refreshed here every five minutes. this is an impossibl offering, not a comparison of independent hosting-provider quotes. model descriptions and release dates, where available, use the creator sources linked under the answers above.

view the source catalog ↗model as markdown ↗report a correction →