Model provider
DeepSeek
Model
deepseek-chat

Fingerprint

245de9d90c396c4ca4128332fad991ff74446cb125f053313703e1fff91ae8dd

Instruction

You label a post for a reader filter. Answer with JSON only.
Topics — choose 1 to 5 from exactly this list, and nothing else:
news, politics, sport, technology, science, health, business, education, culture, religion, environment, travel, food, family, relationships, humour, art, music, film, gaming, community, local, safety, conflict, crime, money, work, housing, transport, weather
Tone — exactly one of: calm, heated, hostile.
Language — the BCP-47 primary subtag of the post, lower case, two to eight letters.
A topic not on the list is a wrong answer, not a new topic.

Answer schema

{
  "type": "object",
  "properties": {
    "topics": {
      "type": "array",
      "items": {
        "type": "string",
        "enum": [
          "news",
          "politics",
          "sport",
          "technology",
          "science",
          "health",
          "business",
          "education",
          "culture",
          "religion",
          "environment",
          "travel",
          "food",
          "family",
          "relationships",
          "humour",
          "art",
          "music",
          "film",
          "gaming",
          "community",
          "local",
          "safety",
          "conflict",
          "crime",
          "money",
          "work",
          "housing",
          "transport",
          "weather"
        ]
      },
      "minItems": 1,
      "maxItems": 5
    },
    "tone": {
      "type": "string",
      "enum": [
        "calm",
        "heated",
        "hostile"
      ]
    },
    "language": {
      "type": "string",
      "pattern": "^[a-zA-Z]{2,8}$"
    }
  },
  "required": [
    "topics",
    "tone",
    "language"
  ],
  "additionalProperties": false
}

Generation parameters

{}

Charter

None.