What is the key capability difference between DeepSeek V3.2 and V3.1?

DeepSeek V3.2 adds combined thinking and tool-use support. Tool calls work in both reasoning and non-reasoning modes, which was a constraint in earlier generation models.

Can DeepSeek V3.2 use tools while in reasoning mode?

Yes. The model supports tool calls in both reasoning and non-reasoning modes.

What is the context window for DeepSeek V3.2?

163.8K tokens, consistent with the V3 model family.

How does DeepSeek V3.2 differ from the Thinking variant?

DeepSeek V3.2 is the general-purpose variant with tool calls in both modes and up to 65.5K tokens. The Thinking variant extends reasoning output to 64K tokens but doesn't support tool use.

Do I need a DeepSeek platform account to use DeepSeek V3.2 through AI Gateway?

No. AI Gateway handles provider authentication. You only need an AI Gateway API key or OIDC token.

When was DeepSeek V3.2 added to AI Gateway?

It became available on December 1, 2025.

DeepSeek

DeepSeek V3.2

DeepSeek V3.2 is DeepSeek's December 1, 2025 model on AI Gateway. It combines tool use with both reasoning and non-reasoning inference modes for agent-style operations.

View API reference

Input and output price: Prices from: Input $0.26, Output $0.38, Per 1M tokens
24h uptime: Loading AI Gateway uptime

Get API key Read the docs

import { streamText } from 'ai'

const result = streamText({
  model: 'deepseek/deepseek-v3.2',
  prompt: 'Why is the sky blue?'
})

Read docs

Overview Playground Providers Uptime Status Throughput Latency API Similar About FAQ

Copy link to headingPlayground

Try out DeepSeek V3.2 by DeepSeek. Usage is billed to your team at API rates. Free users (those who haven't made a payment) get $5 of credits every 30 days.

DeepSeek V3.2

Copy link to headingProviders

Route requests across multiple providers. Copy a provider slug to set your preference. Visit the docs for more info. Using a provider means you agree to their terms, listed under Legal.

Provider

Context	Max Output	Latency	Throughput	Input	Output	Cache	Web Search	Capabilities	ZDR	No Training	Free Tier	Release Date

DeepSeek

Legal:Terms•Privacy

128K

0.8 s

114 tps

$0.28/M

$0.42/M

Read$0.03/M

—

12/01/2025

DeepInfra

Legal:Terms•Privacy

164K

0.7 s

14 tps

$0.26/M

$0.38/M

Read$0.13/M

—

12/01/2025

Novita AI

Legal:Terms•Privacy

164K

66K

1.6 s

30 tps

$0.28/M

$0.42/M

Read$0.13/M

—

12/01/2025

Bedrock

Legal:Terms•Privacy

128K

0.4 s

115 tps

$0.62/M

$1.85/M

—

12/01/2025

Getting started

Call DeepSeek V3.2 through AI Gateway with the AI SDK generateText and streamText functions, or through the OpenAI Chat Completions, OpenAI Responses, and Anthropic Messages APIs by changing the base URL. AI Gateway authenticates the request and routes it to an available provider.

Install the AI SDK (pnpm add ai dotenv), create an API key from the API Keys page, and set it as AI_GATEWAY_API_KEY in your environment. Full setup is covered in the text generation quickstart.

index.ts

import { generateText } from 'ai';
import 'dotenv/config';

async function main() {
  const result = await generateText({
    model: 'deepseek/deepseek-v3.2',
    prompt: 'Why is the sky blue?',
  });

  console.log(result.text);
}

main().catch(console.error);

Top-level parameters

The same DeepSeek V3.2 request in each API format AI Gateway supports.

top-level-params.ts

import { generateText } from 'ai';
import 'dotenv/config';

async function main() {
  const result = await generateText({
    model: 'deepseek/deepseek-v3.2',
    system: 'You are a concise technical assistant.',
    prompt: 'Summarize the tradeoffs between static generation and SSR.',
    maxOutputTokens: 1024,
    temperature: 0.5,
  });

  console.log(result.text);
}

main().catch(console.error);

Standard parameters like prompt, messages, temperature, and tools work as documented in the AI SDK docs. These are the parameters with model-specific behavior.

Parameter	Type	Required	Description
`model`	`string`	Yes	Model ID in the form `creator/model`, e.g. `deepseek/deepseek-v3.2`. AI Gateway routes the request to an available provider.
`maxOutputTokens`	`number`	No	Hard cap on generated tokens. DeepSeek V3.2 supports up to `65,536` output tokens.
`providerOptions`	`Record<string, JSONValue>`	No	AI Gateway routing options under `gateway`, plus any provider-native options under the provider’s own namespace — see the table below.

Input limits

Input	Formats	Sources	Max count	Max size	Limits
Text	—	—	—	—	Prompt and response share the 164K-token context window

Provider options

Set AI Gateway routing options under providerOptions.gateway. For provider-specific options, pass them under the provider’s namespace as documented by the AI SDK.

Learn more in the AI SDK deepseek provider docs.

provider-options.ts

import { generateText } from 'ai';
import 'dotenv/config';

async function main() {
  const result = await generateText({
    model: 'deepseek/deepseek-v3.2',
    prompt: 'Why is the sky blue?',
    providerOptions: {
      gateway: {
        only: ['deepseek', 'deepinfra'],
      },
    },
  });

  console.log(result.text);
}

main().catch(console.error);

These AI Gateway routing options apply to every model. Provider-specific options pass through under the provider’s own namespace (for example providerOptions.anthropic) exactly as documented by the AI SDK.

Parameter	Type	Required	Description
`providerOptions.gateway.only`	`string[]`	No	Restrict routing to these provider slugs. Requests fail over only within the listed providers.
`providerOptions.gateway.order`	`string[]`	No	Preferred provider order. Listed providers are tried first; unlisted providers remain available as fallbacks.
`providerOptions.gateway.sort`	`'cost' \| 'ttft' \| 'tps'`	No	Rank candidate providers by price, time to first token, or tokens per second instead of the default routing order.
`providerOptions.gateway.zeroDataRetention`	`boolean`	No	Route only to providers with a zero-data-retention policy for this model.

Routing across providers

AI Gateway serves the same model through multiple providers and fails over automatically. order expresses a preference while keeping every provider eligible; only is a hard allowlist — if none of the listed providers are available the request fails instead of falling back.

Options under a provider's own namespace (for example providerOptions.anthropic) are forwarded to that provider with the request. Providers ignore option namespaces that don't apply to them, so it is safe to set provider options alongside gateway routing options.

Tool calling

Expose tools the model can call. Define each tool’s inputs with a Zod schema.

tool-calling.ts

import { generateText, tool } from 'ai';
import { z } from 'zod';
import 'dotenv/config';

async function main() {
  const result = await generateText({
    model: 'deepseek/deepseek-v3.2',
    prompt: 'What is the weather in San Francisco?',
    tools: {
      getWeather: tool({
        description: 'Get the current weather for a location',
        inputSchema: z.object({ location: z.string() }),
        execute: async ({ location }) => ({ location, temperatureC: 18 }),
      }),
    },
  });

  console.log(result.text);
}

main().catch(console.error);

Copy link to headingMore models by DeepSeek

Model

Context	Latency	Throughput	Input	Output	Cache	Web Search	Capabilities	Providers	ZDR	No Training	Free Tier	Release Date

deepseek/deepseek-v4-flash-vision-exp

0.6 s

184 tps

$0.22/M

$0.65/M

Read$0.007/M

—

08/21/2026

deepseek/deepseek-v4-pro-0813

0.5 s

123 tps

$0.60/M

$1.90/M

Read$0.02/M

—

08/12/2026

deepseek/deepseek-v4-flash-0731

0.4 s

342 tps

$0.07/M

$0.15/M

Read$0.007/M

—

+16

07/31/2026

deepseek/deepseek-v4-flash

0.3 s

215 tps

$0.08/M

$0.18/M

Read$0.007/M

—

04/23/2026

deepseek/deepseek-v4-pro

0.5 s

155 tps

$0.66/M

$1.98/M

Read$0.02/M

—

04/23/2026

deepseek/deepseek-v3.2-thinking

164K

0.1 s

119 tps

$0.26/M

$0.38/M

Read$0.13/M

—

12/01/2025

Copy link to headingAbout DeepSeek V3.2

DeepSeek V3.2 became available on AI Gateway on December 1, 2025 as the next major iteration of DeepSeek's V3 family. The key capability: the model supports combined thinking and tool use, handling tool calls in both reasoning and non-reasoning modes. This distinguishes it from models where tool use and thinking mode are mutually exclusive, which previously forced developers to choose between the two.

The context window of 163.8K tokens carries over from earlier V3 generation models. Max output is 65.5K tokens in standard chat mode. DeepSeek V3.2 is the general-purpose variant in the V3.2 release, suitable for use cases from chat interfaces to multi-step agent pipelines. For workloads that need maximum reasoning depth and can tolerate higher token consumption, the DeepSeek V3.2 Thinking variant extends reasoning output up to 64,000 tokens but drops tool-use support.

Access through AI Gateway removes the need for a separate provider account. Authentication uses AI Gateway API keys or OIDC tokens, and the AI SDK provides a direct integration path. You can adopt DeepSeek V3.2 without managing DeepSeek platform credentials separately.

Copy link to headingWhat To Consider When Choosing a Provider

Configuration: DeepSeek V3.2 supports tool calls in both reasoning and non-reasoning modes. Test both paths in your integration to confirm your tool schema and response parsing logic handle the output structure from each mode correctly.
Zero Data Retention: Zero Data Retention is available for this model. It is offered on a per-provider and model basis. See the documentation for details.
Authentication: AI Gateway authenticates requests using an API key or OIDC token. You do not need to manage provider credentials directly.

Copy link to headingWhen to Use DeepSeek V3.2

Best for

Combined tool and reasoning: Agentic applications that need both tool calling and reasoning in the same pipeline through one endpoint
General-purpose V3.2 workflows: Chat and instruction-following tasks using the V3.2 generation
Drop-in V3.1 upgrade: Production API integrations using the AI SDK or OpenAI-compatible interfaces
Mixed task pipelines: Tool-augmented completions and reasoning chains served from a single endpoint

Consider alternatives when

Maximum reasoning depth: Use DeepSeek V3.2 Thinking (deepseek-v3.2-thinking) for up to 64K tokens of output when tool use is not needed
Benchmark-level math or code: DeepSeek-R1 remains the dedicated reasoning specialist for math and code reasoning workloads

Copy link to headingConclusion

DeepSeek V3.2 resolves a practical constraint in agent design by supporting tool calls across both reasoning and non-reasoning modes. Available through AI Gateway as of December 1, 2025, it provides a straightforward upgrade path from earlier DeepSeek V3 models.

Copy link to headingFrequently Asked Questions

What is the key capability difference between DeepSeek V3.2 and V3.1?
DeepSeek V3.2 adds combined thinking and tool-use support. Tool calls work in both reasoning and non-reasoning modes, which was a constraint in earlier generation models.
Can DeepSeek V3.2 use tools while in reasoning mode?
Yes. The model supports tool calls in both reasoning and non-reasoning modes.
What is the context window for DeepSeek V3.2?
163.8K tokens, consistent with the V3 model family.
How does DeepSeek V3.2 differ from the Thinking variant?
DeepSeek V3.2 is the general-purpose variant with tool calls in both modes and up to 65.5K tokens. The Thinking variant extends reasoning output to 64K tokens but doesn't support tool use.
Do I need a DeepSeek platform account to use DeepSeek V3.2 through AI Gateway?
No. AI Gateway handles provider authentication. You only need an AI Gateway API key or OIDC token.
When was DeepSeek V3.2 added to AI Gateway?
It became available on December 1, 2025.

Your use is subject to DeepSeek's Terms & Privacy Policies.

Agent Stack

Core Platform

Tools

Learn

Build

Explore

DeepSeek V3.2

Copy link to headingPlayground

Copy link to headingProviders

Copy link to headingUptime24 hours

Copy link to headingThroughput24 hours

Copy link to headingLatency24 hours

Getting started

Top-level parameters

Input limits

Provider options

Routing across providers

Tool calling

Copy link to headingMore models by DeepSeek

Copy link to headingAbout DeepSeek V3.2

Copy link to headingWhat To Consider When Choosing a Provider

Copy link to headingWhen to Use DeepSeek V3.2

Best for

Consider alternatives when

Copy link to headingConclusion

Copy link to headingFrequently Asked Questions