DeepSeek V3.2

DeepSeek V3.2 is DeepSeek's reasoning-first general model built for agents, serving both the chat and reasoner endpoints in non-thinking and thinking modes.

start building →

compare models

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

input · usd / 1m tok

$0.58

output · usd / 1m tok

$1.68

context window

163.84ktokens

modalities · in → out

pricing

base api ratesusd per 1m tokens

input: $0.58
output: $1.68

providers

Azure AI Foundry hosts DeepSeek V3.2. 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-v3-2

creatorDeepSeek
model typeText generation
context window163,840 tokens
maximum outputnot published
input
takes text in, gives text out
outputtext
tool callingnot verified
releaseddec 1, 2025
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-v3-2",
    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-v3-2",
    input="▌",
)
print(resp.output_text)

edit the prompt, then run

benchmarks

intelligence index

16.0#82 of 108 models on the gateway

omniscience-46.9#78 of 84
gpqa75.1%#61 of 80
humanity’s last exam11.2%#68 of 87

Tau278.9%
GPQA75.1%
LiveCodeBench59.3%
AIME 202559.0%
IFBench49.0%
AA-LCR45.7%
Terminal-Bench Hard32.6%
Humanity’s Last Exam11.2%
CritPt0.9%

Intelligence Index16.04
Omniscience Index-46.88
intelligence index (API)16

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 V3.2 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 V3.2?

DeepSeek V3.2 is DeepSeek's reasoning-first general model built for agents, serving both the chat and reasoner endpoints in non-thinking and thinking modes. It introduces DeepSeek Sparse Attention for efficient long-context work and integrates thinking directly into tool use in both modes.

How much does DeepSeek V3.2 cost?

DeepSeek V3.2 costs $0.58 per million input tokens and $1.68 per million output tokens through the Impossibl API. All rates are in USD.

view the pricing table ↑

What is the context length of DeepSeek V3.2?

DeepSeek V3.2 has a 163,840-token context window through the Impossibl API.

What inputs and outputs does DeepSeek V3.2 support?

DeepSeek V3.2 accepts text as input. DeepSeek V3.2 returns text.

How do I use DeepSeek V3.2 through an API?

DeepSeek V3.2 is available through the Impossibl API. Use deepseek/deepseek-v3-2 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 V3.2 released?

DeepSeek released DeepSeek V3.2 on December 1, 2025.

Creator release announcement ↗

this page describes DeepSeek V3.2 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 →