{
  "slug": "esm-c",
  "name": "ESM C",
  "kind": "protein-language-model",
  "publisher": "EvolutionaryScale",
  "tier": "verified",
  "license_code": {
    "status": "permissive",
    "spdx": "MIT",
    "custom_license": false,
    "commercial_use": "yes",
    "derivatives": "yes",
    "redistribution": "yes",
    "license_conditions": [
      "notice-required"
    ],
    "requires_application": false,
    "evidence": [
      {
        "location": "repo-license",
        "excerpt": "Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the “Software”), to deal in the Software without restriction",
        "url": "https://raw.githubusercontent.com/evolutionaryscale/esm/main/LICENSE.md",
        "snapshot_id": "s01",
        "char_start": 64,
        "verified_on": "2026-11-13",
        "observed": "page",
        "captured_by": "pipeline"
      },
      {
        "location": "repo-license",
        "excerpt": "Copyright 2026 Chan Zuckerberg Biohub, Inc.",
        "url": "https://raw.githubusercontent.com/Biohub/esm/main/LICENSE.md",
        "snapshot_id": "s12",
        "char_start": 19,
        "verified_on": "2026-11-13",
        "observed": "page",
        "captured_by": "pipeline"
      }
    ]
  },
  "license_weights": {
    "status": "unverifiable",
    "spdx": "NOASSERTION",
    "custom_license": false,
    "commercial_use": "unverified",
    "derivatives": "unverified",
    "redistribution": "unverified",
    "license_conditions": [],
    "requires_application": false,
    "evidence": [
      {
        "location": "dataset-page",
        "excerpt": "License: other",
        "url": "https://huggingface.co/EvolutionaryScale/esmc-300m-2024-12",
        "snapshot_id": "s08",
        "char_start": 515,
        "verified_on": "2026-11-13",
        "observed": "page",
        "captured_by": "pipeline"
      },
      {
        "location": "dataset-page",
        "excerpt": "\"license:mit\",\"license:other\"",
        "url": "https://huggingface.co/biohub/ESMC-300M",
        "snapshot_id": "s11",
        "char_start": 1343,
        "verified_on": "2026-11-13",
        "observed": "page",
        "captured_by": "pipeline"
      }
    ]
  },
  "last_verified": "2026-11-14",
  "schema_version": "2.0",
  "record_type": "model",
  "official_urls": {
    "repository_or_release": "https://github.com/evolutionaryscale/esm",
    "weights_page": "https://huggingface.co/EvolutionaryScale/esmc-300m-2024-12",
    "paper": null,
    "license_file": "https://raw.githubusercontent.com/evolutionaryscale/esm/main/LICENSE.md"
  },
  "version_history": [
    {
      "version": "unspecified",
      "released_on": "2025-01-16",
      "code_license": null,
      "weights_license": null,
      "changed": "other",
      "note": null,
      "evidence": [
        {
          "location": "dataset-page",
          "excerpt": "ESM C 600M (incl weights)",
          "url": "https://github.com/Biohub/esm/blob/1b3e08025b/LICENSE.md",
          "snapshot_id": "s13",
          "char_start": 1063,
          "verified_on": "2026-11-13",
          "observed": "page",
          "source_work": {
            "title": "Biohub/esm LICENSE.md at commit 1b3e08025b",
            "authors_or_org": "EvolutionaryScale",
            "date": "2025-01-16",
            "date_as_written": "2025-01-16",
            "source_type": "official-page",
            "open_access": true,
            "doi_or_url": "https://github.com/Biohub/esm/blob/1b3e08025b/LICENSE.md"
          }
        }
      ]
    },
    {
      "version": "unspecified",
      "released_on": "2025-01-16",
      "code_license": null,
      "weights_license": null,
      "changed": "other",
      "note": null,
      "evidence": [
        {
          "location": "dataset-page",
          "excerpt": "Just remember to maintain the same license terms and release under the ESM name.",
          "url": "https://github.com/Biohub/esm/blob/1b3e08025b/LICENSE.md",
          "snapshot_id": "s13",
          "char_start": 2382,
          "verified_on": "2026-11-13",
          "observed": "page",
          "source_work": {
            "title": "Biohub/esm LICENSE.md at commit 1b3e08025b",
            "authors_or_org": "EvolutionaryScale",
            "date": "2025-01-16",
            "date_as_written": "2025-01-16",
            "source_type": "official-page",
            "open_access": true,
            "doi_or_url": "https://github.com/Biohub/esm/blob/1b3e08025b/LICENSE.md"
          }
        }
      ]
    },
    {
      "version": "304",
      "released_on": "2026-05-27",
      "code_license": null,
      "weights_license": null,
      "changed": "other",
      "note": null,
      "evidence": [
        {
          "location": "dataset-page",
          "excerpt": "Update esm repo for ESMC paper ( #304 )",
          "url": "https://github.com/Biohub/esm/commit/c94ed8d763",
          "snapshot_id": "s14",
          "char_start": 2547,
          "verified_on": "2026-11-13",
          "observed": "page",
          "source_work": {
            "title": "Biohub/esm commit c94ed8d",
            "authors_or_org": "Chan Zuckerberg Biohub (Biohub/esm)",
            "date": "2026-05-27",
            "date_as_written": "2026-05-27",
            "source_type": "official-page",
            "open_access": true,
            "doi_or_url": "https://github.com/Biohub/esm/commit/c94ed8d763"
          }
        }
      ]
    }
  ],
  "output_restrictions": [
    {
      "kind": "other",
      "applies_to": "all",
      "statement": {
        "text": {
          "en": "The model card's out-of-scope section prohibits any use barred by the Acceptable Use Policy.",
          "zh": "模型卡“不适用范围”禁止可接受使用政策所禁止的任何用途。"
        },
        "evidence": [
          {
            "location": "dataset-page",
            "excerpt": "Any use that is prohibited by the Acceptable Use Policy",
            "url": "https://huggingface.co/biohub/ESMC-300M",
            "snapshot_id": "s11",
            "char_start": 18658,
            "verified_on": "2026-11-13",
            "observed": "page",
            "source_work": {
              "title": "ESMC model card — Out-of-Scope or Unauthorized Use Cases",
              "authors_or_org": "Chan Zuckerberg Biohub (Biohub/esm)",
              "date": null,
              "date_as_written": null,
              "source_type": "official-page",
              "open_access": true,
              "doi_or_url": "https://huggingface.co/biohub/ESMC-300M"
            }
          }
        ]
      }
    }
  ],
  "training_data": [
    {
      "resource": "UniRef, MGnify, and the Joint Genome Institute (JGI)",
      "slug": "uniprot",
      "role": "pretraining",
      "cutoff_date": null,
      "cutoff_stated": false,
      "statement": {
        "text": {
          "en": "ESM C is trained on protein sequences from UniRef, MGnify, and the Joint Genome Institute (JGI).",
          "zh": "ESM C 使用来自 UniRef、MGnify 和 Joint Genome Institute (JGI) 的蛋白质序列进行训练。"
        },
        "evidence": [
          {
            "location": "dataset-page",
            "excerpt": "ESM C is trained on protein sequences from UniRef, MGnify, and the Joint Genome Institute (JGI).",
            "url": "https://www.evolutionaryscale.ai/blog/esm-cambrian",
            "snapshot_id": "s02",
            "char_start": 7326,
            "verified_on": "2026-11-13",
            "observed": "page",
            "captured_by": "pipeline"
          }
        ]
      }
    },
    {
      "resource": "UniRef",
      "slug": "uniprot",
      "role": "other",
      "cutoff_date": null,
      "cutoff_stated": false,
      "statement": {
        "text": {
          "en": "ESMC was trained on UniRef, MGnify and JGI sequences clustered at 70% identity (83M, 372M and 2B clusters respectively).",
          "zh": "ESMC 训练数据来自 UniRef、MGnify 与 JGI，按 70% 序列一致性聚类（分别 8300 万、3.72 亿、20 亿簇）。"
        },
        "evidence": [
          {
            "location": "dataset-page",
            "excerpt": "ESMC was trained on protein sequences from UniRef, MGnify, and the Joint Genome Institute (JGI).",
            "url": "https://huggingface.co/biohub/ESMC-300M",
            "snapshot_id": "s11",
            "char_start": 9479,
            "verified_on": "2026-11-13",
            "observed": "page",
            "source_work": {
              "title": "ESMC model card — Training Data",
              "authors_or_org": "Chan Zuckerberg Biohub (Biohub/esm)",
              "date": null,
              "date_as_written": null,
              "source_type": "official-page",
              "open_access": true,
              "doi_or_url": "https://huggingface.co/biohub/ESMC-300M"
            }
          }
        ]
      }
    }
  ],
  "benchmark_overlap": [],
  "access": [
    {
      "method": "open-weights-download",
      "url": "https://huggingface.co/EvolutionaryScale/esmc-300m-2024-12",
      "terms_differ_from_weights_license": false,
      "own_terms": null
    },
    {
      "method": "hosted-api",
      "url": "https://biohub.ai",
      "terms_differ_from_weights_license": false,
      "own_terms": null
    },
    {
      "method": "third-party-hosted",
      "url": "https://github.com/Biohub/esm/blob/1b3e08025b/LICENSE.md",
      "terms_differ_from_weights_license": false,
      "own_terms": {
        "text": {
          "en": "In the 2025 licence, a free non-commercial inference API was offered via Forge (EvolutionaryScale).",
          "zh": "2025 版许可中，通过 Forge 提供免费的非商业推理 API。"
        },
        "evidence": [
          {
            "location": "dataset-page",
            "excerpt": "Free non-commercial inference API",
            "url": "https://github.com/Biohub/esm/blob/1b3e08025b/LICENSE.md",
            "snapshot_id": "s13",
            "char_start": 2468,
            "verified_on": "2026-11-13",
            "observed": "page",
            "source_work": {
              "title": "Biohub/esm LICENSE.md at commit 1b3e08025b",
              "authors_or_org": "EvolutionaryScale",
              "date": "2025-01-16",
              "date_as_written": "2025-01-16",
              "source_type": "official-page",
              "open_access": true,
              "doi_or_url": "https://github.com/Biohub/esm/blob/1b3e08025b/LICENSE.md"
            }
          }
        ]
      }
    }
  ],
  "conflicts": [],
  "hardware_notes": [
    {
      "text": {
        "en": "The model card only recommends a GPU-enabled PyTorch environment, with transformer_engine and xformers optional.",
        "zh": "模型卡仅推荐带 GPU 的 PyTorch 环境，transformer_engine、xformers 为可选。"
      },
      "evidence": [
        {
          "location": "dataset-page",
          "excerpt": "PyTorch environment with GPU support recommended.",
          "url": "https://huggingface.co/biohub/ESMC-300M",
          "snapshot_id": "s11",
          "char_start": 9333,
          "verified_on": "2026-11-13",
          "observed": "page",
          "source_work": {
            "title": "ESMC model card — System Requirements",
            "authors_or_org": "Chan Zuckerberg Biohub (Biohub/esm)",
            "date": null,
            "date_as_written": null,
            "source_type": "official-page",
            "open_access": true,
            "doi_or_url": "https://huggingface.co/biohub/ESMC-300M"
          }
        }
      ]
    }
  ],
  "known_defects": [
    {
      "kind": "bias",
      "statement": {
        "text": {
          "en": "The model card lists dataset bias: over- or under-representation of taxa and protein families in public databases can bias outputs.",
          "zh": "模型卡列出数据集偏差：公共数据库中物种、蛋白家族的过多或过少代表会使输出产生偏差。"
        },
        "evidence": [
          {
            "location": "dataset-page",
            "excerpt": "Over- or under-representation of taxa, protein families, or ecological niches in public sequence and structure databases influences generalization and can bias outputs.",
            "url": "https://huggingface.co/biohub/ESMC-300M",
            "snapshot_id": "s11",
            "char_start": 18119,
            "verified_on": "2026-11-13",
            "observed": "page",
            "source_work": {
              "title": "ESMC model card — Potential Biases",
              "authors_or_org": "Chan Zuckerberg Biohub (Biohub/esm)",
              "date": null,
              "date_as_written": null,
              "source_type": "official-page",
              "open_access": true,
              "doi_or_url": "https://huggingface.co/biohub/ESMC-300M"
            }
          }
        ]
      },
      "source_kind": "official-note",
      "scope": null,
      "attribution": {
        "who": "Chan Zuckerberg Biohub (Biohub/esm)",
        "when": "date not stated (page read 2026-11-13)",
        "source_type": "official-page",
        "title": "ESMC model card — Potential Biases"
      },
      "correction_entrance": "Report an error through the publisher or the authors: https://huggingface.co/biohub/ESMC-300M"
    }
  ],
  "fit": {
    "suited_for": [
      {
        "text": {
          "en": "For token or sequence classification, the classifier head is not pretrained and is meant to be fine-tuned for the downstream task.",
          "zh": "做 token 或序列分类时，分类头未经预训练，需针对下游任务微调。"
        },
        "evidence": [
          {
            "location": "dataset-page",
            "excerpt": "the classifier head is not pretrained but instead meant to be fine-tuned for your downstream task.",
            "url": "https://huggingface.co/biohub/ESMC-300M",
            "snapshot_id": "s11",
            "char_start": 17017,
            "verified_on": "2026-11-13",
            "observed": "page",
            "source_work": {
              "title": "ESMC model card — Usage",
              "authors_or_org": "Chan Zuckerberg Biohub (Biohub/esm)",
              "date": null,
              "date_as_written": null,
              "source_type": "official-page",
              "open_access": true,
              "doi_or_url": "https://huggingface.co/biohub/ESMC-300M"
            }
          }
        ]
      }
    ],
    "not_suited_for": []
  },
  "gaps": [
    {
      "field": "benchmark_overlap",
      "note": "未找到有出处的说明 / no sourced statement found"
    }
  ],
  "completeness": {
    "core_complete": true
  },
  "changelog": [
    {
      "date": "2026-11-14",
      "change": {
        "en": "Added by the automated pipeline from stored page snapshots; every statement carries a verbatim quote.",
        "zh": "由自动流水线根据已存页面快照添加，每条陈述带逐字引文。"
      },
      "by": "pipeline"
    }
  ],
  "_links": {
    "text": "Continue in this database. This record is one of 109 processed and verified records (datasets, data resources and open models). To find others, call `https://mcp.mdexa.com/v1/datasets?q=<words>` or connect the MCP server at `https://mcp.mdexa.com/mcp`; no key is needed for 20 calls per minute. Rows are under 1 KB; an empty result means \"not listed\". For a list of items, one batch check (free account) returns a report. Guide: /agents.md. Similar records: ESM3 open (esm3-sm-open-v1) (https://mdexa.com/models/esm3-open/); AlphaFold 3 (https://mdexa.com/models/alphafold3/); Boltz (Boltz-1 and Boltz-2) (https://mdexa.com/models/boltz/).",
    "records": 109,
    "search": {
      "template": "https://mcp.mdexa.com/v1/datasets?q={words}",
      "example": "https://mcp.mdexa.com/v1/datasets?q=binding%20affinity"
    },
    "mcp": "https://mcp.mdexa.com/mcp",
    "anonymous_calls_per_minute": 20,
    "guide": "https://mdexa.com/agents.md",
    "faq": "https://mdexa.com/developers/faq/",
    "similar": [
      {
        "name": "ESM3 open (esm3-sm-open-v1)",
        "url": "https://mdexa.com/models/esm3-open/",
        "json": "https://mdexa.com/models/esm3-open.json"
      },
      {
        "name": "AlphaFold 3",
        "url": "https://mdexa.com/models/alphafold3/",
        "json": "https://mdexa.com/models/alphafold3.json"
      },
      {
        "name": "Boltz (Boltz-1 and Boltz-2)",
        "url": "https://mdexa.com/models/boltz/",
        "json": "https://mdexa.com/models/boltz.json"
      }
    ]
  }
}