> ## Documentation Index
> Fetch the complete documentation index at: https://veniceai-feat-models-redesign.mintlify.site/llms.txt
> Use this file to discover all available pages before exploring further.

# DeepSeek V3.2 API

> DeepSeek V3.2 API on Venice: 160K context, $0.33 input and $0.48 output per 1M tokens. Private, with zero data retention. Pricing, specs and code examples.

export const HubMount = ({view, children, ...props}) => {
  const [hub, setHub] = useState(null);
  const [failed, setFailed] = useState(false);
  useEffect(() => {
    let alive = true;
    const w = window;
    if (!w.__veniceModelHub) {
      const urls = w.location.hostname === 'localhost' ? ['http://localhost:3333/data/model-hub.bundle.json', '/data/model-hub.bundle.json'] : ['/data/model-hub.bundle.json'];
      const load = i => fetch(urls[i], {
        cache: 'no-cache'
      }).then(res => {
        if (!res.ok) throw new Error(`bundle ${res.status}`);
        return res.json();
      }).catch(err => i + 1 < urls.length ? load(i + 1) : Promise.reject(err));
      const Frag = <></>.type;
      const h = (type, props, ...kids) => {
        const T = type;
        const {key, ...rest} = props || ({});
        if (!kids.length) return <T key={key} {...rest} />;
        if (kids.length === 1) return <T key={key} {...rest}>{kids[0]}</T>;
        return <T key={key} {...rest}>{kids.map((kid, i) => <Frag key={i}>{kid}</Frag>)}</T>;
      };
      w.__veniceModelHub = load(0).then(bundle => new Function(`return (${bundle.code})`)()({
        h,
        Fragment: Frag,
        useState,
        useEffect,
        useRef,
        useMemo,
        useCallback
      }));
    }
    w.__veniceModelHub.then(instance => {
      if (alive) setHub(instance);
    }).catch(() => {
      w.__veniceModelHub = null;
      if (alive) setFailed(true);
    });
    return () => {
      alive = false;
    };
  }, []);
  const View = hub ? hub[view] : null;
  if (View) return <View {...props}>{children}</View>;
  if (failed) {
    return <div className="vx-mount is-failed">
        <p className="vx-mount-note">The interactive model catalog could not load. The full data is below.</p>
        {children}
      </div>;
  }
  return <div className="vx-mount" aria-busy="true">
      <div className="vx-mount-skeleton" aria-hidden="true"><span /><span /><span /></div>
      <div className="vx-mount-source">{children}</div>
    </div>;
};

<HubMount view="ModelPage" data={{"family":{"slug":"deepseek-v3-2","name":"DeepSeek V3.2","modality":"text","task":"chat","provider":"deepseek","description":"DeepSeek-V3.2 is an efficient large language model with DeepSeek Sparse Attention (DSA) for long contexts. It features strong reasoning and tool-use skills, achieving top results on the 2025 IMO and IOI.","primary":"deepseek-v3.2","variants":["deepseek-v3.2"],"created":1764806400,"updated":1764806400,"privacy":["private"],"openWeights":true},"models":[{"id":"deepseek-v3.2","name":"DeepSeek V3.2","type":"text","modality":"text","task":"chat","variant":"standard","provider":"deepseek","created":1764806400,"description":"DeepSeek-V3.2 is an efficient large language model with DeepSeek Sparse Attention (DSA) for long contexts. It features strong reasoning and tool-use skills, achieving top results on the 2025 IMO and IOI.","source":"https://huggingface.co/deepseek-ai/DeepSeek-V3.2","privacy":"private","openWeights":true,"text":{"context":160000,"maxOutput":32768,"reasoning":{"supported":true,"effort":["none","low","medium","high"],"defaultEffort":"low"},"caps":{"tools":true,"structured":true,"webSearch":true}},"pricing":{"input":0.33,"output":0.48,"cacheRead":0.16,"blended":0.3675},"headline":{"value":0.3675,"unit":"per 1M tokens","basis":"blended"},"endpoints":[{"id":"chat","method":"POST","path":"/chat/completions","name":"Chat Completions","status":"stable","recommended":true},{"id":"responses","method":"POST","path":"/responses","name":"Responses","status":"alpha"}],"family":"deepseek-v3-2"}],"related":{"similar":[{"slug":"venice-uncensored-1-2","name":"Venice Uncensored 1.2","provider":"mistral","modality":"text","privacy":["private"],"created":1775001600,"headline":{"value":0.375,"unit":"per 1M tokens","basis":"blended"},"variants":1},{"slug":"gpt-4o-mini","name":"GPT-4o Mini","provider":"openai","modality":"text","privacy":["anonymized"],"created":1772236800,"headline":{"value":0.328125,"unit":"per 1M tokens","basis":"blended"},"variants":1},{"slug":"gemma-4-26b-a4b-uncensored","name":"Gemma 4 26B A4B Uncensored","provider":"google","modality":"text","privacy":["e2ee"],"created":1779580800,"headline":{"value":0.3625,"unit":"per 1M tokens","basis":"blended"},"variants":1},{"slug":"qwen3-vl-30b-a3b","name":"Qwen3 VL 30B A3B","provider":"alibaba","modality":"text","privacy":["e2ee"],"created":1773792000,"headline":{"value":0.4125,"unit":"per 1M tokens","basis":"blended"},"variants":1}],"versions":[]},"providers":{"deepseek":{"slug":"deepseek","name":"DeepSeek","logo":"/images/icons/models/deepseek.svg"},"mistral":{"slug":"mistral","name":"Mistral AI","logo":"/images/icons/models/mistral.svg"},"openai":{"slug":"openai","name":"OpenAI","logo":"/images/icons/models/openai.svg"},"google":{"slug":"google","name":"Google","logo":"/images/icons/models/google.svg"},"alibaba":{"slug":"alibaba","name":"Alibaba Qwen","logo":"/images/icons/models/qwen.svg"}},"faq":[{"q":"How much does the DeepSeek V3.2 API cost?","a":"$0.33 per 1M input tokens and $0.48 per 1M output tokens, with cached input at $0.16 per 1M. Prices are in USD and can be paid in DIEM at parity."},{"q":"What is the DeepSeek V3.2 model ID?","a":"Use `deepseek-v3.2` as the `model` parameter."},{"q":"Is the DeepSeek V3.2 API private?","a":"DeepSeek V3.2 is private: requests run on infrastructure Venice controls with zero data retention, and prompts and outputs are never stored or used for training."},{"q":"What is the context window of DeepSeek V3.2?","a":"160K tokens of context, with up to 32K output tokens per response."},{"q":"What does DeepSeek V3.2 support?","a":"DeepSeek V3.2 supports function calling, structured outputs, reasoning, web search and prompt caching. Reasoning effort is adjustable with `reasoning_effort`: none, low, medium and high (default low)."},{"q":"Which endpoint does the DeepSeek V3.2 API use?","a":"Call `POST /chat/completions`. `/responses` (Alpha) is also supported."}]}} />

<div className="vx-static">
  <Accordion title="Plain-text specification">
    # DeepSeek V3.2 API

    DeepSeek V3.2 is a large language model by DeepSeek, available on the Venice API as `deepseek-v3.2`. It runs privately, with zero data retention.

    DeepSeek-V3.2 is an efficient large language model with DeepSeek Sparse Attention (DSA) for long contexts. It features strong reasoning and tool-use skills, achieving top results on the 2025 IMO and IOI.

    ## DeepSeek V3.2 API pricing

    | Model ID | Variant | Privacy | Price |
    | - | - | - | - |
    | `deepseek-v3.2` | Standard | Private | $0.33 input / $0.48 output per 1M tokens |

    ## DeepSeek V3.2 specifications

    | Spec | Value |
    | - | - |
    | Provider | DeepSeek |
    | Released | Dec 4, 2025 |
    | Privacy | Private |
    | Open weights | Yes |
    | Context window | 160K tokens |
    | Max output | 32K tokens |
    | Reasoning effort | none, low, medium, high |

    ## How to use the DeepSeek V3.2 API

    Send requests to `POST https://api.venice.ai/api/v1/chat/completions` with `"model": "deepseek-v3.2"` and your API key.

    ```bash theme={null}
    curl https://api.venice.ai/api/v1/chat/completions \
      -H "Authorization: Bearer $VENICE_API_KEY" \
      -H "Content-Type: application/json" \
      -d '{
        "model": "deepseek-v3.2",
        "messages": [{ "role": "user", "content": "Explain TEE attestation in two sentences." }],
        "reasoning_effort": "low"
      }'
    ```

    ## DeepSeek V3.2 API FAQ

    ### How much does the DeepSeek V3.2 API cost?

    $0.33 per 1M input tokens and $0.48 per 1M output tokens, with cached input at \$0.16 per 1M. Prices are in USD and can be paid in DIEM at parity.

    ### What is the DeepSeek V3.2 model ID?

    Use `deepseek-v3.2` as the `model` parameter.

    ### Is the DeepSeek V3.2 API private?

    DeepSeek V3.2 is private: requests run on infrastructure Venice controls with zero data retention, and prompts and outputs are never stored or used for training.

    ### What is the context window of DeepSeek V3.2?

    160K tokens of context, with up to 32K output tokens per response.

    ### What does DeepSeek V3.2 support?

    DeepSeek V3.2 supports function calling, structured outputs, reasoning, web search and prompt caching. Reasoning effort is adjustable with `reasoning_effort`: none, low, medium and high (default low).

    ### Which endpoint does the DeepSeek V3.2 API use?

    Call `POST /chat/completions`. `/responses` (Alpha) is also supported.

    ## Related models

    * [Venice Uncensored 1.2 API](/models/venice-uncensored-1-2): $0.20 input / $0.90 output per 1M tokens
    * [GPT-4o Mini API](/models/gpt-4o-mini): $0.19 input / $0.75 output per 1M tokens
    * [Gemma 4 26B A4B Uncensored API](/models/gemma-4-26b-a4b-uncensored): $0.19 input / $0.88 output per 1M tokens
    * [Qwen3 VL 30B A3B API](/models/qwen3-vl-30b-a3b): $0.25 input / $0.90 output per 1M tokens
  </Accordion>
</div>


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.