{
 "schema_version": "1.0",
 "dataset": "deepseek-models",
 "verified_at": "2026-11-04",
 "models": [
  {
   "id": "deepseek-v4.1-flash-1017",
   "display_name": "DeepSeek Flash",
   "family": "flash",
   "positioning": "Fast, cheap, multimodal. Default choice for classification, extraction, translation and everyday chat.",
   "context_tokens": 1000000,
   "max_output_tokens": 384000,
   "vision_input": true,
   "max_concurrency": 2500,
   "compare_to": {
    "deepseek-v4-pro": {
     "input_price_multiple": 4.4,
     "output_price_multiple": 3.3
    }
   },
   "typical_use": [
    "classification",
    "extraction",
    "translation",
    "summarisation",
    "chat",
    "document Q&A",
    "vision OCR"
   ]
  },
  {
   "id": "deepseek-v4-pro",
   "display_name": "DeepSeek V4 Pro",
   "family": "pro",
   "positioning": "Reasoning-heavy tasks. Use when a wrong answer costs more than the tokens.",
   "context_tokens": 1000000,
   "max_output_tokens": 384000,
   "vision_input": false,
   "max_concurrency": 500,
   "typical_use": [
    "code review",
    "debugging",
    "complex analysis",
    "long-form writing",
    "professional judgement"
   ]
  }
 ]
}