{
  "id": "all-minilm-l6-v2",
  "identity": {
    "family": "sentence-transformers",
    "display_name": "all-MiniLM-L6-v2",
    "hub_id": "sentence-transformers/all-MiniLM-L6-v2",
    "api_name": null,
    "revision": null,
    "variants": [],
    "license": "Apache-2.0",
    "commercial_use": true,
    "commercial_use_note": "Apache-2.0 weights; no use restrictions.",
    "release_date": null,
    "paper_url": "https://arxiv.org/abs/1908.10084",
    "repo_url": "https://github.com/UKPLab/sentence-transformers"
  },
  "task": {
    "category": "embedding",
    "modalities_in": [
      "text"
    ],
    "modalities_out": [
      "vector"
    ],
    "best_at": "Small, fast English sentence vectors on CPU (22M params, 384 dims) for short text, dedup and clustering.",
    "worst_at": "Non-English text, long passages (truncated at 256 tokens here), and hard retrieval where E5-large wins."
  },
  "io": {
    "input": {
      "text": {
        "max_tokens": 256,
        "truncation": true,
        "language": "English"
      },
      "prompts": "none"
    },
    "output": {
      "vector": {
        "dims": 384,
        "distance": "cosine",
        "type": "dense"
      }
    }
  },
  "mixpeek_wiring": {
    "extractors": [
      {
        "extractor": "text_extractor@v1",
        "selected_by": "embedding_model=all_minilm_l6_v2_v1"
      }
    ],
    "feature_uris": [
      "mixpeek://text_extractor@v1/all_minilm_l6_v2_v1"
    ],
    "vector_indexes": [
      "text_extractor_v1_embedding"
    ],
    "retriever_stages": [
      "feature_search"
    ],
    "composes_with": [
      "text_extractor with embedding_model=all_minilm_l6_v2_v1 (dimensions fixed at namespace creation)"
    ]
  },
  "run": {
    "status": "served",
    "status_reason": "Ray Serve app in the inference registry; text_extractor can select it per run. The registry marks it a small fast fallback, E5 is the default.",
    "serve_apps": [
      "sentence_transformers__all_MiniLM_L6_v2"
    ],
    "runtime": "transformers AutoModel + AutoTokenizer (BertModel), mean pooling",
    "weights_format": "safetensors",
    "trust_remote_code": false,
    "size_on_disk_mb": 91,
    "params_millions": 22.7,
    "preprocessing": "Tokenize with truncation at 256 tokens.",
    "postprocessing": "Mean pool over the attention mask, L2-normalize (384 dims).",
    "serving_shape": "both",
    "batch_size": null,
    "serve_batch": null,
    "scale_to_zero": {
      "min_replicas": 0,
      "max_replicas": 4,
      "idle_timeout_s": 300,
      "memory_gb": 0.5,
      "num_cpus": 0.25,
      "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": "No committed throughput or cost measurement for this model."
  },
  "evidence": {
    "unmeasured": false,
    "items": [
      {
        "dataset": "LongMemEval subset, 60 questions (hybrid_flat, run outside Mixpeek)",
        "metric": "QA accuracy, SYSTEM number: local MiniLM dense + rank_bm25 RRF retrieval, Gemini reader; not Mixpeek's Serve app",
        "k": null,
        "score": 0.783,
        "baseline": null,
        "baseline_score": null,
        "held_out": false,
        "date": "2026-09-27",
        "source": "research/agent-memory/RESULTS.md"
      }
    ]
  },
  "refs": [
    "hf://sentence-transformers/all-MiniLM-L6-v2@1110a243fdf4706b3f48f1d95db1a4f5529b4d41"
  ],
  "formats": [
    {
      "ref": "hf://sentence-transformers/all-MiniLM-L6-v2@1110a243fdf4706b3f48f1d95db1a4f5529b4d41",
      "format": "safetensors",
      "file": "model.safetensors",
      "loader": "transformers / open_clip from_pretrained; Mixpeek Hub import (api/models/huggingface.py)",
      "supported_today": true
    },
    {
      "ref": "hf://sentence-transformers/all-MiniLM-L6-v2@1110a243fdf4706b3f48f1d95db1a4f5529b4d41",
      "format": "onnx",
      "file": "9 files, e.g. onnx/model.onnx",
      "loader": "onnxruntime; Mixpeek model upload (api/models/loadability.py, ONNX IR <= 11)",
      "supported_today": true
    },
    {
      "ref": "hf://sentence-transformers/all-MiniLM-L6-v2@1110a243fdf4706b3f48f1d95db1a4f5529b4d41",
      "format": "pytorch_bin",
      "file": "3 files, e.g. openvino/openvino_model.bin",
      "loader": "torch.load (pickle); Mixpeek refuses pickle weights (shared/security/torch_pickle.py)",
      "supported_today": false
    },
    {
      "ref": "hf://sentence-transformers/all-MiniLM-L6-v2@1110a243fdf4706b3f48f1d95db1a4f5529b4d41",
      "format": "tensorflow",
      "file": "tf_model.h5",
      "loader": "TensorFlow; not in Mixpeek's runtime",
      "supported_today": false
    }
  ],
  "use_cases": [
    "text-search"
  ],
  "recommended": {
    "value": false,
    "for": [],
    "reason": null
  },
  "demos": [],
  "seo": {
    "title": "all-MiniLM-L6-v2: embedding model",
    "description": "all-MiniLM-L6-v2 is a text embedding model (Apache-2.0). Served on Mixpeek, scaling to zero when idle.",
    "keywords": [
      "all-MiniLM-L6-v2",
      "sentence-transformers",
      "embedding",
      "sentence-transformers/all-MiniLM-L6-v2",
      "text-search"
    ],
    "faqs": [
      {
        "q": "What is all-MiniLM-L6-v2 best at?",
        "a": "Small, fast English sentence vectors on CPU (22M params, 384 dims) for short text, dedup and clustering."
      },
      {
        "q": "What are the limits of all-MiniLM-L6-v2?",
        "a": "Non-English text, long passages (truncated at 256 tokens here), and hard retrieval where E5-large wins."
      },
      {
        "q": "Can all-MiniLM-L6-v2 be used commercially?",
        "a": "Yes. Apache-2.0 weights; no use restrictions."
      },
      {
        "q": "How does all-MiniLM-L6-v2 run on Mixpeek?",
        "a": "Ray Serve app in the inference registry; text_extractor can select it per run. The registry marks it a small fast fallback, E5 is the default."
      }
    ]
  },
  "last_verified": "2026-11-05",
  "sources": [
    "https://huggingface.co/sentence-transformers/all-MiniLM-L6-v2",
    "https://huggingface.co/api/models/sentence-transformers/all-MiniLM-L6-v2",
    "https://arxiv.org/abs/1908.10084",
    "server/engine/inference/sentence_transformers/all_minilm_l6_v2/services.py",
    "server/shared/models/embeddings/registry.py",
    "server/engine/extractors/text_extractor/v1/definition.py"
  ],
  "registry": {
    "sentence_transformers__all_MiniLM_L6_v2": {
      "deployment": "sentence_transformers__all_MiniLM_L6_v2",
      "service_id": "sentence-transformers/all-MiniLM-L6-v2",
      "routes": "engine/inference/sentence_transformers/all_minilm_l6_v2/routes.py",
      "route_prefix": "/sentence_transformers/all_minilm_l6_v2/v1",
      "enabled": true,
      "description": "All-MiniLM-L6-v2 sentence transformer. Generates 384-dimensional dense embeddings for text. Lightweight 22M-param CPU-only model for staging/testing.",
      "capabilities": {
        "category": "embedding",
        "modality_input": [
          "text"
        ],
        "modality_output": "vector",
        "supports_batch": true,
        "reranking_mode": null
      },
      "vector_config": {
        "dimensions": 384,
        "distance": "cosine",
        "vector_type": "dense"
      },
      "serve": {
        "min_replicas": 0,
        "max_replicas": 4,
        "idle_timeout_s": 300,
        "target_ongoing_requests": 3,
        "num_cpus": 0.25,
        "num_gpus": 0,
        "memory_gb": 0.5,
        "pinnable": false,
        "placement": "worker",
        "enabled": true
      },
      "interactive_twin": false,
      "interactive_serve": null,
      "schemas": {
        "input": "shared.inference.sentence_transformers.all_minilm_l6_v2.models.InferenceRequest",
        "output": "shared.inference.sentence_transformers.all_minilm_l6_v2.models.InferenceResponse",
        "config": "shared.inference.sentence_transformers.all_minilm_l6_v2.models.InferenceConfigs"
      }
    }
  },
  "templates": [],
  "usage": {
    "namespaces": null,
    "collections": null,
    "uncounted": true,
    "templates": 0,
    "counted_at": null,
    "source": null
  },
  "facets": {
    "status": "served",
    "category": "embedding",
    "family": "sentence-transformers",
    "license": "Apache-2.0",
    "commercial_use": true,
    "modalities_in": [
      "text"
    ],
    "modalities_out": [
      "vector"
    ],
    "use_cases": [
      "text-search"
    ],
    "recommended": false,
    "release_date": null,
    "params_millions": 22.7,
    "size_on_disk_mb": 91,
    "vector_dims": 384,
    "vector_distance": "cosine",
    "weights_format": "safetensors",
    "formats": [
      "onnx",
      "pytorch_bin",
      "safetensors",
      "tensorflow"
    ],
    "loadable_today": true,
    "trust_remote_code": false,
    "cpu_fallback": true,
    "num_gpus": 0,
    "memory_gb": 0.5,
    "min_replicas": 0,
    "serving_shape": "both",
    "served": true,
    "ref_schemes": [
      "hf"
    ],
    "has_evidence": true,
    "cost_measured": false,
    "template_count": 0,
    "usage_namespaces": null,
    "usage_collections": null,
    "usage_uncounted": true,
    "has_demo": false,
    "rank_score": 4
  },
  "rank_score": 4
}