chore(models): auto-add new and deprecate stale DeepSeek models

[github-actions]

Auto-generated PR from .github/workflows/chore-update-provider-models.yml.

Scan results for DeepSeek

{
  "provider": "DeepSeek",
  "apiUrl": "https://openrouter.ai/api/v1/models",
  "providerApiQueried": true,
  "providerApiUrl": "https://api.deepseek.com/v1/models",
  "apiModels": [
    "deepseek-v4-flash",
    "deepseek-v4-pro"
  ],
  "openrouterModels": [
    "deepseek-chat",
    "deepseek-chat-v3-0324",
    "deepseek-chat-v3.1",
    "deepseek-r1",
    "deepseek-r1-0528",
    "deepseek-r1-distill-llama-70b",
    "deepseek-r1-distill-qwen-32b",
    "deepseek-v3.1-terminus",
    "deepseek-v3.2",
    "deepseek-v3.2-exp",
    "deepseek-v3.2-speciale",
    "deepseek-v4-flash",
    "deepseek-v4-flash:free",
    "deepseek-v4-pro"
  ],
  "providerApiModels": [
    "deepseek-v4-flash",
    "deepseek-v4-pro"
  ],
  "sourceModels": [
    "deepseek-chat",
    "deepseek-reasoner",
    "deepseek-v4-flash",
    "deepseek-v4-pro"
  ],
  "newModels": [],
  "deprecatedModels": [
    "deepseek-chat",
    "deepseek-reasoner"
  ],
  "unchangedModels": [
    "deepseek-v4-flash",
    "deepseek-v4-pro"
  ],
  "fileUpdated": true,
  "validation": {
    "success": true,
    "errors": [],
    "missingDefaultCapabilities": [],
    "pendingCapabilityModels": [],
    "realtimeModels": []
  }
}

Actions taken:

• New models discovered via OpenRouter are auto-inserted with capabilities mapped from architecture.modalities and supported_parameters.
• Disappeared models (or models with expiration_date < 1 year) are marked Deprecated = true.
• Existing model Capabilities and ContextLimit are refreshed from OpenRouter metadata.
• Rank values are auto-computed from OpenRouter data: newer models rank higher, and within the same creation period cheaper output pricing ranks higher.

Requires manual review:

• Verify Verified flags for newly added models.