DeepSeek: DeepSeek V3.2 Speciale

Description

DeepSeek-V3.2-Speciale is a high-compute variant of DeepSeek-V3.2 optimized for maximum reasoning and agentic performance. It builds on DeepSeek Sparse Attention (DSA) for efficient long-context processing, then scales post-training reinforcement learning to push capability beyond the base model. Reported evaluations place Speciale ahead of GPT-5 on difficult reasoning workloads, with proficiency comparable to Gemini-3.0-Pro, while retaining strong coding and tool-use reliability. Like V3.2, it benefits from a large-scale agentic task synthesis pipeline that improves compliance and generalization in interactive environments.

Raw fields snapshot

{
  "id": "deepseek/deepseek-v3.2-speciale",
  "name": "DeepSeek: DeepSeek V3.2 Speciale",
  "description": "DeepSeek-V3.2-Speciale is a high-compute variant of DeepSeek-V3.2 optimized for maximum reasoning and agentic performance. It builds on DeepSeek Sparse Attention (DSA) for efficient long-context processing, then scales post-training reinforcement learning to push capability beyond the base model. Reported evaluations place Speciale ahead of GPT-5 on difficult reasoning workloads, with proficiency comparable to Gemini-3.0-Pro, while retaining strong coding and tool-use reliability. Like V3.2, it benefits from a large-scale agentic task synthesis pipeline that improves compliance and generalization in interactive environments.",
  "created": 1764594837,
  "canonical_slug": "deepseek/deepseek-v3.2-speciale-20251201",
  "hugging_face_id": "deepseek-ai/DeepSeek-V3.2-Speciale",
  "source_type": "openrouter_only",
  "context_length": 163840,
  "max_completion_tokens": 65536,
  "is_moderated": false,
  "architecture": {
    "modality": "text->text",
    "input_modalities": [
      "text"
    ],
    "output_modalities": [
      "text"
    ],
    "tokenizer": "DeepSeek",
    "instruct_type": null
  },
  "input_modalities": [
    "text"
  ],
  "output_modalities": [
    "text"
  ],
  "modality": "text->text",
  "tokenizer": "DeepSeek",
  "instruct_type": null,
  "supported_parameters": [
    "include_reasoning",
    "max_tokens",
    "reasoning",
    "response_format",
    "structured_outputs",
    "temperature",
    "top_p"
  ],
  "default_parameters": {
    "temperature": 1,
    "top_p": 0.95,
    "frequency_penalty": null
  },
  "per_request_limits": null,
  "top_provider": {
    "context_length": 163840,
    "max_completion_tokens": 65536,
    "is_moderated": false
  },
  "pricing": {
    "prompt": "0.0000004",
    "completion": "0.0000012",
    "input_cache_read": "0.0000002"
  },
  "PPM": {
    "prompt": 0.4,
    "completion": 1.2,
    "input_cache_read": 0.2
  },
  "openrouter_raw": {
    "id": "deepseek/deepseek-v3.2-speciale",
    "canonical_slug": "deepseek/deepseek-v3.2-speciale-20251201",
    "hugging_face_id": "deepseek-ai/DeepSeek-V3.2-Speciale",
    "name": "DeepSeek: DeepSeek V3.2 Speciale",
    "created": 1764594837,
    "description": "DeepSeek-V3.2-Speciale is a high-compute variant of DeepSeek-V3.2 optimized for maximum reasoning and agentic performance. It builds on DeepSeek Sparse Attention (DSA) for efficient long-context processing, then scales post-training reinforcement learning to push capability beyond the base model. Reported evaluations place Speciale ahead of GPT-5 on difficult reasoning workloads, with proficiency comparable to Gemini-3.0-Pro, while retaining strong coding and tool-use reliability. Like V3.2, it benefits from a large-scale agentic task synthesis pipeline that improves compliance and generalization in interactive environments.",
    "context_length": 163840,
    "architecture": {
      "modality": "text->text",
      "input_modalities": [
        "text"
      ],
      "output_modalities": [
        "text"
      ],
      "tokenizer": "DeepSeek",
      "instruct_type": null
    },
    "pricing": {
      "prompt": "0.0000004",
      "completion": "0.0000012",
      "input_cache_read": "0.0000002"
    },
    "top_provider": {
      "context_length": 163840,
      "max_completion_tokens": 65536,
      "is_moderated": false
    },
    "per_request_limits": null,
    "supported_parameters": [
      "include_reasoning",
      "max_tokens",
      "reasoning",
      "response_format",
      "structured_outputs",
      "temperature",
      "top_p"
    ],
    "default_parameters": {
      "temperature": 1,
      "top_p": 0.95,
      "frequency_penalty": null
    },
    "expiration_date": null
  }
}