impulse

V3.2

Open weights
API id deepseek.v3.2
Input / 1M
$0.62
at Amazon Bedrock
Output / 1M
$1.85
at Amazon Bedrock
Context window
156K
tokens
Max output
8K
tokens per reply
Hosts
1
12 listings
Released
-
date not published

Pricing

List prices at Amazon Bedrock, in US dollars per million tokens.

TierInput / 1MOutput / 1M
Standard$0.62$1.85
Batch
50% off
$0.31$0.925
Flex
50% off
$0.31$0.925
Priority
75% more
$1.08$3.24

What it costs a month

At Amazon Bedrock list prices, 30 days.

  • Chat assistant5,000 requests a day, 2K in / 0.5K out$325
  • RAG over documents2,000 requests a day, 8K in / 0.6K out$364
  • Coding agent1,000 requests a day, 40K in / 2K out$855
  • Overnight batch20,000 requests a day, 4K in / 0.4K out, batch$966
Use your own numbers

Where to run it

1 host serve V3.2. The comparison is on blended price, 3 parts input to 1 part output.

HostIn / out per 1M
deepseek.v3.2
$0.62 / $1.85
Show 11 more listings: regions, snapshots and variants
Amazon Bedrock
us-east-1
deepseek.v3.2
$0.62 / $1.85
us-east-2
deepseek.v3.2
$0.62 / $1.85
us-west-2
deepseek.v3.2
$0.62 / $1.85
Mantle
deepseek.v3.2
$0.62 / $1.85
us
us.deepseek.v3.2
$0.62 / $1.85
ap-northeast-1
deepseek.v3.2
$0.74 / $2.22
+20% vs headline
ap-south-1
deepseek.v3.2
$0.74 / $2.22
+20% vs headline
ap-southeast-3
deepseek.v3.2
$0.74 / $2.22
+20% vs headline
eu-north-1
deepseek.v3.2
$0.74 / $2.22
+20% vs headline
sa-east-1
deepseek.v3.2
$0.74 / $2.22
+20% vs headline
eu
eu.deepseek.v3.2
$0.74 / $2.22
+20% vs headline

Capabilities

Core

  • Streaming
  • Function calling
  • JSON mode
  • Structured outputs

Multimodal

  • Vision
  • Image input
  • Audio
  • Image generation
  • Video

Advanced

  • Reasoning
  • Tool calling
  • MCP
  • Prompt caching
  • Batch API
  • Embeddings
Takes in
Text
Produces
Text
Features
Native structured output
Open weights
License: DeepSeek License (MIT code)
Runtimes: SGLang, TGI, vLLM

Cheaper alternatives

Current models that cost less and keep most of what this one can do: tools, vision, reasoning, structured output, caching and enough context.

Same model, cheaper host

No host sells it for less than Amazon Bedrock.

Smaller DeepSeek models

No cheaper DeepSeek model keeps its key capabilities.

Other frontier labs

  • Gemini 3.5 Flash Lite
    Google · on Google AI Studio
    $0.30 / $2.50
    Has every key capability
    8% less
  • GPT-5.4 Nano
    OpenAI · on OpenAI
    $0.20 / $1.25
    Has every key capability
    50% less
  • Claude Haiku 5.5
    Anthropic · on Anthropic
    $0.10 / $0.50
    Has every key capability
    78% less
  • GPT-6 Luna
    OpenAI · on OpenAI
    $0.10 / $0.50
    Has every key capability
    78% less

Open-weight picks

  • GLM 5.3
    Zhipu AI · on CLF AI Gateway
    $0.56 / $1.76
    Has every key capability
    7% less
  • DeepSeek V4 Pro
    DeepSeek · on CLF AI Gateway
    $0.528 / $1.58
    Has every key capability
    15% less
  • Kimi K2.7 Code
    Moonshot AI · on CLF AI Gateway
    $0.38 / $1.60
    Has every key capability
    26% less
  • Inkling Small
    Independent · on DeepInfra
    $0.45 / $1.20
    Has every key capability
    31% less

Model line

Newer and older versions, at each lab's own price.

  1. UndatedV3.2Latest$0.62 / $1.85
  2. UndatedV3.1$0.58 / $1.68
  3. UndatedV3$0.58 / $1.68

Call it

Using Amazon Bedrock's API.

Python
from litellm import completion

response = completion(
    model="bedrock/deepseek.v3.2",
    messages=[{"role": "user", "content": "Hello"}],
)
print(response.choices[0].message.content)

Quick answers

How much does V3.2 cost?

On Amazon Bedrock, V3.2 costs $0.62 per million input tokens and $1.85 per million output tokens.

Which APIs offer V3.2?

V3.2 runs on 1 API: Amazon Bedrock.

What is V3.2's context window?

156K tokens, with up to 8K tokens in one reply.

Is V3.2 being retired?

No retirement date is scheduled for it on any API we track.

Is there a newer version of V3.2?

No. V3.2 is the newest model in its line.

What are cheaper alternatives to V3.2?

Models that cost less and keep most of what it can do: Gemini 3.5 Flash Lite on Google AI Studio (8% less), GPT-5.4 Nano on OpenAI (50% less) and GLM 5.3 on CLF AI Gateway (7% less). Savings are on a mix of 3 input tokens to 1 output token.

What is V3.2's API model id?

Use deepseek.v3.2 with Amazon Bedrock.

Sources

Raw catalog entry for deepseek.v3.2 on Amazon Bedrock
{
  "input_cost_per_token": 6.2e-7,
  "input_cost_per_token_batches": 3.1e-7,
  "input_cost_per_token_flex": 3.1e-7,
  "input_cost_per_token_priority": 0.000001085,
  "litellm_provider": "bedrock_converse",
  "max_input_tokens": 156000,
  "max_output_tokens": 8000,
  "mode": "chat",
  "output_cost_per_token": 0.00000185,
  "output_cost_per_token_batches": 9.25e-7,
  "output_cost_per_token_flex": 9.25e-7,
  "output_cost_per_token_priority": 0.0000032375,
  "supports_function_calling": true,
  "supports_reasoning": true,
  "supports_native_structured_output": true,
  "supports_tool_choice": true,
  "source": "https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-deepseek-deepseek-v3-2.md",
  "supports_audio_input": false,
  "supports_response_schema": true,
  "supports_vision": false
}