# DeepSeek V3.2

> DeepSeek V3.2 is an efficiency-focused reasoning model from DeepSeek using sparse attention, with a 32,768-token context window and strong coding and agentic tool-use. Served free through SambaNova Cloud's fast RDU inference and SiliconFlow.


**ID**: `deepseek/deepseek-v3.2`  
**Creator**: DeepSeek  
**Category**: text  
**Context**: 33K tokens  
**Status**: Disabled (since 2026-08-27) — no longer routable or listed; kept for history/usage reference.  
**Reason**: The remaining platform host is out of credit, and the other platform route is already disabled. Remaining routes require a user-owned key, so the model is unlisted rather than advertised as a live hosted chat model.  
**Input**: $0.27/M  
**Output**: $0.40/M  
**Released**: 2025-12-01  
**Web page**: https://anyrouter.dev/model/deepseek/deepseek-v3.2

**Input modalities**: text  
**Output modalities**: text  

**Capabilities**: chat, reasoning, streaming, coding, function-calling

**Aliases**: `deepseek/DeepSeek-V3.2`, `deepseek-ai/DeepSeek-V3.2`

## Usage

**Endpoint**: `POST https://anyrouter.dev/api/v1/chat/completions`

```bash
curl https://anyrouter.dev/api/v1/chat/completions \
  -H "Authorization: Bearer $ANYROUTER_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "model": "deepseek/deepseek-v3.2",
  "messages": [
    {
      "role": "user",
      "content": "Say hi in 3 words."
    }
  ]
}'
```

## Providers

| Provider | Input | Output |
| --- | --- | --- |
| sambanova | $0 | $0 |
| sambanova-byok | $0 | $0 |
| siliconflow | $0 | $0 |
| siliconflow-byok | $0 | $0 |
| cline-byok | $0 | $0 |

## Benchmarks

- **GPQA Diamond**: 0.8
- **HLE**: 0.1
- **τ²-Bench Telecom**: 0.8
- **SciCode**: 0.4

## Source

- Upstream docs: https://api-docs.deepseek.com