{
  "id": "gemini-3-5-flash-lite",
  "identity": {
    "family": "gemini",
    "display_name": "Gemini 3.5 Flash-Lite",
    "hub_id": null,
    "api_name": "gemini-3.5-flash-lite",
    "revision": null,
    "variants": [],
    "license": "Google Gemini API / Vertex AI terms of service",
    "commercial_use": true,
    "commercial_use_note": "Paid Google API under Google's terms; tenants can bring their own key.",
    "release_date": null,
    "paper_url": null,
    "repo_url": null
  },
  "task": {
    "category": "generation",
    "modalities_in": [
      "text",
      "image",
      "video",
      "document"
    ],
    "modalities_out": [
      "text",
      "structured"
    ],
    "best_at": "The platform's default LLM (DEFAULT_GEMINI_MODEL) for cheap per-document work in retriever stages (enrich, filter, summarize) and for video-segment description and OCR in multimodal_extractor v1.",
    "worst_at": "Multi-step reasoning and long structured extraction, where the code picks gemini-2.5-flash or escalates to gemini-2.5-pro; per-document cost grows with candidate count in llm_filter and llm_enrich."
  },
  "io": {
    "input": {
      "text": "prompt plus document fields",
      "image": "URLs or inline data",
      "video": "segments (multimodal_extractor v1)"
    },
    "output": {
      "text": "free text, or JSON when response_shape / response_schema is set"
    }
  },
  "mixpeek_wiring": {
    "extractors": [
      {
        "extractor": "multimodal_extractor@v1",
        "selected_by": "fixed (video description and OCR when run_video_description or run_ocr)"
      },
      {
        "extractor": "universal_extractor@v1",
        "selected_by": "fixed (GENERATION_MODEL)"
      },
      {
        "extractor": "scrolling_text_extractor@v1",
        "selected_by": "fixed (direct Gemini OCR fallback, _OCR_FALLBACK_MODEL)"
      }
    ],
    "feature_uris": [],
    "vector_indexes": [],
    "retriever_stages": [
      "llm_enrich",
      "llm_filter",
      "summarize",
      "agent_search",
      "agentic_enrich",
      "rerank"
    ],
    "composes_with": [
      "feature_search -> llm_filter or llm_enrich (default model_name) -> summarize",
      "rerank with inference_name mixpeek__decision and decision_model=gemini-3.5-flash-lite"
    ]
  },
  "run": {
    "status": "api",
    "status_reason": "Provider API. Retriever stages and extractors call it through CentralLLMService or google-genai directly; universal__api routes any gemini-* model for POST /v1/inference.",
    "serve_apps": [
      "universal__api"
    ],
    "runtime": "Gemini developer API or Vertex AI (clients.providers.google.gemini_backend, shared.models.llm GoogleProvider)",
    "weights_format": "api",
    "trust_remote_code": false,
    "size_on_disk_mb": null,
    "params_millions": null,
    "preprocessing": "Prompt templates per stage; media passed as URLs or inline data.",
    "postprocessing": "Text, or JSON parsed against the requested schema.",
    "serving_shape": "both",
    "batch_size": null,
    "serve_batch": null,
    "scale_to_zero": {
      "min_replicas": 0,
      "max_replicas": 6,
      "idle_timeout_s": 600,
      "memory_gb": null,
      "num_cpus": 0.1,
      "num_gpus": 0,
      "cold_start_s": {
        "warm_node_s": null,
        "new_node_s": null,
        "source": null
      },
      "cpu_fallback": true
    }
  },
  "cost": {
    "unmeasured": true,
    "measured": null,
    "note": "Billed per token by Google; no committed throughput and cost measurement."
  },
  "evidence": {
    "unmeasured": true,
    "items": []
  },
  "refs": [
    "api://google/gemini-3.5-flash-lite"
  ],
  "formats": [
    {
      "ref": "api://google/gemini-3.5-flash-lite",
      "format": "api",
      "file": null,
      "loader": "provider SDK, called by model name",
      "supported_today": true
    }
  ],
  "use_cases": [
    "video-understanding",
    "tagging",
    "captioning",
    "ocr",
    "brand-detection"
  ],
  "recommended": {
    "value": true,
    "for": [
      "tagging",
      "captioning",
      "ocr"
    ],
    "reason": "Platform default Gemini model since 2026-10-10 (shared/models/llm/gemini_models.py)."
  },
  "demos": [],
  "seo": {
    "title": "Gemini 3.5 Flash-Lite: language model",
    "description": "Gemini 3.5 Flash-Lite is a language model. Called on Mixpeek through the provider API.",
    "keywords": [
      "Gemini 3.5 Flash-Lite",
      "gemini",
      "language",
      "gemini-3.5-flash-lite",
      "video-understanding",
      "tagging",
      "captioning",
      "ocr",
      "brand-detection"
    ],
    "faqs": [
      {
        "q": "What is Gemini 3.5 Flash-Lite best at?",
        "a": "The platform's default LLM (DEFAULT_GEMINI_MODEL) for cheap per-document work in retriever stages (enrich, filter, summarize) and for video-segment description and OCR in multimodal_extractor v1."
      },
      {
        "q": "What are the limits of Gemini 3.5 Flash-Lite?",
        "a": "Multi-step reasoning and long structured extraction, where the code picks gemini-2.5-flash or escalates to gemini-2.5-pro; per-document cost grows with candidate count in llm_filter and llm_enrich."
      },
      {
        "q": "Can Gemini 3.5 Flash-Lite be used commercially?",
        "a": "Yes. Paid Google API under Google's terms; tenants can bring their own key."
      },
      {
        "q": "How does Gemini 3.5 Flash-Lite run on Mixpeek?",
        "a": "Provider API. Retriever stages and extractors call it through CentralLLMService or google-genai directly; universal__api routes any gemini-* model for POST /v1/inference."
      }
    ]
  },
  "last_verified": "2026-11-05",
  "sources": [
    "https://ai.google.dev/gemini-api/docs/models/gemini-3.5-flash-lite",
    "server/shared/models/llm/gemini_models.py",
    "server/shared/retrievers/stages/enrich/llm/models.py",
    "server/shared/retrievers/stages/filter/llm/models.py",
    "server/shared/retrievers/stages/reduce/summarize/models.py",
    "server/shared/retrievers/stages/enrich/agentic/models.py",
    "server/engine/extractors/multimodal_extractor/v1/pipeline.py",
    "server/engine/extractors/universal_extractor/v1/processors/gemini_client.py",
    "server/engine/extractors/scrolling_text_extractor/v1/processors/scrolling_text_batch.py",
    "server/engine/inference/llm/routes.py"
  ],
  "registry": {
    "universal__api": {
      "deployment": "universal__api",
      "service_id": "universal/api",
      "routes": "engine/inference/llm/routes.py",
      "route_prefix": "/universal/api",
      "enabled": true,
      "description": "Universal inference endpoint routing to multiple providers and modalities. Supports: OpenAI (GPT, embeddings, Whisper), Google (Gemini, Vertex embeddings), Anthropic (Claude). Routes based on model parameter.",
      "capabilities": {
        "category": "generation",
        "modality_input": [
          "text"
        ],
        "modality_output": "text",
        "supports_batch": true,
        "reranking_mode": null
      },
      "vector_config": null,
      "serve": {
        "min_replicas": 0,
        "max_replicas": 6,
        "idle_timeout_s": 600,
        "target_ongoing_requests": 25,
        "num_cpus": 0.1,
        "num_gpus": 0,
        "memory_gb": null,
        "pinnable": false,
        "placement": "worker",
        "enabled": true
      },
      "interactive_twin": false,
      "interactive_serve": null,
      "schemas": {
        "input": null,
        "output": null,
        "config": null
      }
    }
  },
  "templates": [],
  "usage": {
    "namespaces": null,
    "collections": null,
    "uncounted": true,
    "templates": 0,
    "counted_at": null,
    "source": null
  },
  "facets": {
    "status": "api",
    "category": "generation",
    "family": "gemini",
    "license": "Google Gemini API / Vertex AI terms of service",
    "commercial_use": true,
    "modalities_in": [
      "text",
      "image",
      "video",
      "document"
    ],
    "modalities_out": [
      "text",
      "structured"
    ],
    "use_cases": [
      "video-understanding",
      "tagging",
      "captioning",
      "ocr",
      "brand-detection"
    ],
    "recommended": true,
    "release_date": null,
    "params_millions": null,
    "size_on_disk_mb": null,
    "vector_dims": null,
    "vector_distance": null,
    "weights_format": "api",
    "formats": [
      "api"
    ],
    "loadable_today": true,
    "trust_remote_code": false,
    "cpu_fallback": true,
    "num_gpus": 0,
    "memory_gb": null,
    "min_replicas": 0,
    "serving_shape": "both",
    "served": true,
    "ref_schemes": [
      "api"
    ],
    "has_evidence": false,
    "cost_measured": false,
    "template_count": 0,
    "usage_namespaces": null,
    "usage_collections": null,
    "usage_uncounted": true,
    "has_demo": false,
    "rank_score": 3
  },
  "rank_score": 3
}