{
  "slug": "bentoml",
  "name": "BentoML",
  "company": "BentoML",
  "website": "https://www.bentoml.com/",
  "category": "platforms",
  "business_model": "open",
  "availability": "unknown",
  "launched": null,
  "tags": [
    "infer",
    "developer-tools"
  ],
  "updated": "2026-11-03",
  "sources": [
    {
      "title": "BentoML — official product",
      "url": "https://www.bentoml.com/"
    },
    {
      "title": "BentoML — official documentation and terms",
      "url": "https://docs.bentoml.com/"
    }
  ],
  "selection": {
    "tasks": [
      "model-hosting"
    ],
    "form_factor": "library",
    "platforms": [
      "unknown"
    ],
    "deployment": [
      "self-hosted"
    ],
    "offline": "unknown",
    "regions": [
      "unknown"
    ],
    "api": "yes",
    "license": "unknown",
    "commercial_use": "unknown",
    "languages": [
      "unknown"
    ],
    "billing": {
      "models": [
        "open-source",
        "enterprise-quote"
      ],
      "free_allowance": "yes",
      "requires_subscription": "no"
    }
  },
  "selection_evidence": {
    "tasks": {
      "status": "editorial_inference",
      "note": "Editorial classification of the documented features; not a product-performance test."
    },
    "form_factor": {
      "status": "editorial_inference",
      "note": "Editorial classification of the documented features; not a product-performance test."
    },
    "platforms": {
      "status": "unknown",
      "note": "This product-specific field was not established in this entry’s source review."
    },
    "deployment": {
      "status": "editorial_inference",
      "note": "Editorial classification of the documented features; not a product-performance test."
    },
    "offline": {
      "status": "unknown",
      "note": "This product-specific field was not established in this entry’s source review."
    },
    "regions": {
      "status": "unknown",
      "note": "This product-specific field was not established in this entry’s source review."
    },
    "api": {
      "status": "editorial_inference",
      "note": "Editorial classification of the documented features; not a product-performance test."
    },
    "license": {
      "status": "unknown",
      "note": "This product-specific field was not established in this entry’s source review."
    },
    "commercial_use": {
      "status": "unknown",
      "note": "This product-specific field was not established in this entry’s source review."
    },
    "languages": {
      "status": "unknown",
      "note": "This product-specific field was not established in this entry’s source review."
    },
    "billing": {
      "status": "editorial_inference",
      "note": "Editorial classification of the documented features; not a product-performance test."
    }
  },
  "facts": [
    {
      "id": "bentoml.capability.primary",
      "field": "capability",
      "value": "Build inference APIs and compose multiple models",
      "status": "verified",
      "evidence_type": "official_page",
      "checked": "2026-11-03",
      "sources": [
        "https://docs.bentoml.com/"
      ]
    }
  ],
  "docs_url": "https://docs.bentoml.com/",
  "description": "Python serving framework for packaging AI models and custom inference pipelines into deployable services.",
  "overview": "Typical uses include Serving AI models to applications; Deployment of production inference endpoints. Documented capabilities: Build inference APIs and compose multiple models; Package dependencies and models into reproducible containers; Adaptive batching and asynchronous task queues.",
  "features": [
    "Build inference APIs and compose multiple models",
    "Package dependencies and models into reproducible containers",
    "Adaptive batching and asynchronous task queues"
  ],
  "use_cases": [
    "Serving AI models to applications",
    "Deployment of production inference endpoints"
  ],
  "pricing_notes": "Self-hosting the open-source project and buying the hosted platform are separate choices. The hosted service has its own plans and usage limits; self-hosting still requires infrastructure and any model-provider charges.",
  "faq": [
    {
      "q": "Where does BentoML fit in a workflow?",
      "a": "Python serving framework for packaging AI models and custom inference pipelines into deployable services. Build inference APIs and compose multiple models."
    },
    {
      "q": "What costs and limits should I compare before choosing BentoML?",
      "a": "Self-hosting the open-source project and buying the hosted platform are separate choices. The hosted service has its own plans and usage limits; self-hosting still requires infrastructure and any model-provider charges."
    }
  ],
  "i18n": {
    "zh": {
      "description": "将AI模型及自定义推理流水线封装为可部署服务的Python框架。",
      "overview": "典型用途包括向应用提供AI模型、生产推理端点部署。文档列明的能力包括：构建推理API及组合多个模型；将依赖和模型打包为可复现容器；自适应批处理及异步任务队列。",
      "features": [
        "构建推理API及组合多个模型",
        "将依赖和模型打包为可复现容器",
        "自适应批处理及异步任务队列"
      ],
      "use_cases": [
        "向应用提供AI模型",
        "生产推理端点部署"
      ],
      "pricing_notes": "开源项目自托管与购买托管平台是两种选择。托管服务有独立套餐及用量限制，自托管仍需承担基础设施和模型提供商费用。",
      "faq": [
        {
          "q": "BentoML适合工作流中的哪个环节？",
          "a": "将AI模型及自定义推理流水线封装为可部署服务的Python框架。 构建推理API及组合多个模型."
        },
        {
          "q": "选择BentoML前应比较哪些费用和限制？",
          "a": "开源项目自托管与购买托管平台是两种选择。托管服务有独立套餐及用量限制，自托管仍需承担基础设施和模型提供商费用。"
        }
      ],
      "selection_evidence": {
        "tasks": {
          "note": "依据所记录功能作出的编辑归类，非产品性能实测。"
        },
        "form_factor": {
          "note": "依据所记录功能作出的编辑归类，非产品性能实测。"
        },
        "platforms": {
          "note": "本条目的来源核查尚未确认这一产品字段。"
        },
        "deployment": {
          "note": "依据所记录功能作出的编辑归类，非产品性能实测。"
        },
        "offline": {
          "note": "本条目的来源核查尚未确认这一产品字段。"
        },
        "regions": {
          "note": "本条目的来源核查尚未确认这一产品字段。"
        },
        "api": {
          "note": "依据所记录功能作出的编辑归类，非产品性能实测。"
        },
        "license": {
          "note": "本条目的来源核查尚未确认这一产品字段。"
        },
        "commercial_use": {
          "note": "本条目的来源核查尚未确认这一产品字段。"
        },
        "languages": {
          "note": "本条目的来源核查尚未确认这一产品字段。"
        },
        "billing": {
          "note": "依据所记录功能作出的编辑归类，非产品性能实测。"
        }
      },
      "facts": {
        "bentoml.capability.primary": {
          "value": "构建推理API及组合多个模型"
        }
      }
    },
    "es": {
      "description": "Framework Python para empaquetar modelos y pipelines de inferencia en servicios.",
      "overview": "Usos habituales: Servicio de modelos para aplicaciones; Despliegue de endpoints de inferencia. Funciones documentadas: API de inferencia y composición de modelos; Dependencias y modelos en contenedores reproducibles; Lotes adaptativos y colas asíncronas.",
      "features": [
        "API de inferencia y composición de modelos",
        "Dependencias y modelos en contenedores reproducibles",
        "Lotes adaptativos y colas asíncronas"
      ],
      "use_cases": [
        "Servicio de modelos para aplicaciones",
        "Despliegue de endpoints de inferencia"
      ],
      "pricing_notes": "Autoalojar el proyecto abierto y contratar la plataforma alojada son opciones distintas. El servicio tiene planes y límites propios; el autoalojamiento requiere infraestructura y cargos de modelos.",
      "faq": [
        {
          "q": "¿En qué parte del flujo encaja BentoML?",
          "a": "Framework Python para empaquetar modelos y pipelines de inferencia en servicios. API de inferencia y composición de modelos."
        },
        {
          "q": "¿Qué costes y límites debo comparar antes de elegir BentoML?",
          "a": "Autoalojar el proyecto abierto y contratar la plataforma alojada son opciones distintas. El servicio tiene planes y límites propios; el autoalojamiento requiere infraestructura y cargos de modelos."
        }
      ],
      "selection_evidence": {
        "tasks": {
          "note": "Clasificación editorial de las funciones documentadas; no es una prueba de rendimiento."
        },
        "form_factor": {
          "note": "Clasificación editorial de las funciones documentadas; no es una prueba de rendimiento."
        },
        "platforms": {
          "note": "La revisión de fuentes de esta entrada no estableció este dato del producto."
        },
        "deployment": {
          "note": "Clasificación editorial de las funciones documentadas; no es una prueba de rendimiento."
        },
        "offline": {
          "note": "La revisión de fuentes de esta entrada no estableció este dato del producto."
        },
        "regions": {
          "note": "La revisión de fuentes de esta entrada no estableció este dato del producto."
        },
        "api": {
          "note": "Clasificación editorial de las funciones documentadas; no es una prueba de rendimiento."
        },
        "license": {
          "note": "La revisión de fuentes de esta entrada no estableció este dato del producto."
        },
        "commercial_use": {
          "note": "La revisión de fuentes de esta entrada no estableció este dato del producto."
        },
        "languages": {
          "note": "La revisión de fuentes de esta entrada no estableció este dato del producto."
        },
        "billing": {
          "note": "Clasificación editorial de las funciones documentadas; no es una prueba de rendimiento."
        }
      },
      "facts": {
        "bentoml.capability.primary": {
          "value": "API de inferencia y composición de modelos"
        }
      }
    },
    "fr": {
      "description": "Framework Python pour empaqueter modèles et inférence en services déployables.",
      "overview": "Usages courants : Mise à disposition de modèles IA; Déploiement de points d’inférence en production. Fonctions documentées : API d’inférence et composition de modèles; Dépendances et modèles en conteneurs reproductibles; Lots adaptatifs et files asynchrones.",
      "features": [
        "API d’inférence et composition de modèles",
        "Dépendances et modèles en conteneurs reproductibles",
        "Lots adaptatifs et files asynchrones"
      ],
      "use_cases": [
        "Mise à disposition de modèles IA",
        "Déploiement de points d’inférence en production"
      ],
      "pricing_notes": "L’auto-hébergement du projet et le service hébergé sont deux options distinctes. Le service dispose de ses propres forfaits et quotas ; l’infrastructure et les modèles restent à financer en auto-hébergement.",
      "faq": [
        {
          "q": "Quelle place occupe BentoML dans un parcours de travail ?",
          "a": "Framework Python pour empaqueter modèles et inférence en services déployables. API d’inférence et composition de modèles."
        },
        {
          "q": "Quels coûts et limites comparer avant de choisir BentoML ?",
          "a": "L’auto-hébergement du projet et le service hébergé sont deux options distinctes. Le service dispose de ses propres forfaits et quotas ; l’infrastructure et les modèles restent à financer en auto-hébergement."
        }
      ],
      "selection_evidence": {
        "tasks": {
          "note": "Classement éditorial des fonctions documentées ; aucun test de performance."
        },
        "form_factor": {
          "note": "Classement éditorial des fonctions documentées ; aucun test de performance."
        },
        "platforms": {
          "note": "L’examen des sources de cette fiche n’a pas établi ce point du produit."
        },
        "deployment": {
          "note": "Classement éditorial des fonctions documentées ; aucun test de performance."
        },
        "offline": {
          "note": "L’examen des sources de cette fiche n’a pas établi ce point du produit."
        },
        "regions": {
          "note": "L’examen des sources de cette fiche n’a pas établi ce point du produit."
        },
        "api": {
          "note": "Classement éditorial des fonctions documentées ; aucun test de performance."
        },
        "license": {
          "note": "L’examen des sources de cette fiche n’a pas établi ce point du produit."
        },
        "commercial_use": {
          "note": "L’examen des sources de cette fiche n’a pas établi ce point du produit."
        },
        "languages": {
          "note": "L’examen des sources de cette fiche n’a pas établi ce point du produit."
        },
        "billing": {
          "note": "Classement éditorial des fonctions documentées ; aucun test de performance."
        }
      },
      "facts": {
        "bentoml.capability.primary": {
          "value": "API d’inférence et composition de modèles"
        }
      }
    },
    "ja": {
      "description": "AIモデルと独自推論パイプラインを展開可能なサービスにまとめるPythonフレームワーク。",
      "overview": "主な用途はアプリへのAIモデル提供、本番推論エンドポイントの展開です。資料に記載された機能：推論APIの作成と複数モデルの組合せ；依存とモデルを再現可能なコンテナへ梱包；適応バッチ処理と非同期タスクキュー。",
      "features": [
        "推論APIの作成と複数モデルの組合せ",
        "依存とモデルを再現可能なコンテナへ梱包",
        "適応バッチ処理と非同期タスクキュー"
      ],
      "use_cases": [
        "アプリへのAIモデル提供",
        "本番推論エンドポイントの展開"
      ],
      "pricing_notes": "オープンソース版の自前運用とホスト型プラットフォームは別の選択肢です。ホスト型には独自のプランと上限があり、自前運用でも基盤とモデル利用料が必要です。",
      "faq": [
        {
          "q": "BentoMLはワークフローのどの部分に向いていますか？",
          "a": "AIモデルと独自推論パイプラインを展開可能なサービスにまとめるPythonフレームワーク。 推論APIの作成と複数モデルの組合せ."
        },
        {
          "q": "BentoMLを選ぶ前に比較すべき費用や制限は何ですか？",
          "a": "オープンソース版の自前運用とホスト型プラットフォームは別の選択肢です。ホスト型には独自のプランと上限があり、自前運用でも基盤とモデル利用料が必要です。"
        }
      ],
      "selection_evidence": {
        "tasks": {
          "note": "記録された機能に基づく編集上の分類であり、性能の実測ではありません。"
        },
        "form_factor": {
          "note": "記録された機能に基づく編集上の分類であり、性能の実測ではありません。"
        },
        "platforms": {
          "note": "この項目の資料確認では、この製品固有の情報は未確認です。"
        },
        "deployment": {
          "note": "記録された機能に基づく編集上の分類であり、性能の実測ではありません。"
        },
        "offline": {
          "note": "この項目の資料確認では、この製品固有の情報は未確認です。"
        },
        "regions": {
          "note": "この項目の資料確認では、この製品固有の情報は未確認です。"
        },
        "api": {
          "note": "記録された機能に基づく編集上の分類であり、性能の実測ではありません。"
        },
        "license": {
          "note": "この項目の資料確認では、この製品固有の情報は未確認です。"
        },
        "commercial_use": {
          "note": "この項目の資料確認では、この製品固有の情報は未確認です。"
        },
        "languages": {
          "note": "この項目の資料確認では、この製品固有の情報は未確認です。"
        },
        "billing": {
          "note": "記録された機能に基づく編集上の分類であり、性能の実測ではありません。"
        }
      },
      "facts": {
        "bentoml.capability.primary": {
          "value": "推論APIの作成と複数モデルの組合せ"
        }
      }
    },
    "de": {
      "description": "Python-Servingframework für Modelle und eigene Inferenzpipelines als Dienste.",
      "overview": "Typische Anwendungen: Bereitstellung von KI-Modellen; Deployment produktiver Inferenzendpunkte. Dokumentierte Funktionen: Inferenz-APIs und mehrere Modelle kombinieren; Abhängigkeiten und Modelle in reproduzierbare Container packen; Adaptive Stapelverarbeitung und asynchrone Warteschlangen.",
      "features": [
        "Inferenz-APIs und mehrere Modelle kombinieren",
        "Abhängigkeiten und Modelle in reproduzierbare Container packen",
        "Adaptive Stapelverarbeitung und asynchrone Warteschlangen"
      ],
      "use_cases": [
        "Bereitstellung von KI-Modellen",
        "Deployment produktiver Inferenzendpunkte"
      ],
      "pricing_notes": "Selbsthosting des offenen Projekts und die gehostete Plattform sind getrennte Angebote. Der Dienst hat eigene Tarife und Limits; beim Selbsthosting fallen Infrastruktur- und Modellkosten an.",
      "faq": [
        {
          "q": "Wo passt BentoML in den Arbeitsablauf?",
          "a": "Python-Servingframework für Modelle und eigene Inferenzpipelines als Dienste. Inferenz-APIs und mehrere Modelle kombinieren."
        },
        {
          "q": "Welche Kosten und Limits sollte ich vor der Wahl von BentoML vergleichen?",
          "a": "Selbsthosting des offenen Projekts und die gehostete Plattform sind getrennte Angebote. Der Dienst hat eigene Tarife und Limits; beim Selbsthosting fallen Infrastruktur- und Modellkosten an."
        }
      ],
      "selection_evidence": {
        "tasks": {
          "note": "Redaktionelle Einordnung dokumentierter Funktionen; kein Leistungstest."
        },
        "form_factor": {
          "note": "Redaktionelle Einordnung dokumentierter Funktionen; kein Leistungstest."
        },
        "platforms": {
          "note": "Die Quellenprüfung dieses Eintrags hat dieses Produktfeld nicht belegt."
        },
        "deployment": {
          "note": "Redaktionelle Einordnung dokumentierter Funktionen; kein Leistungstest."
        },
        "offline": {
          "note": "Die Quellenprüfung dieses Eintrags hat dieses Produktfeld nicht belegt."
        },
        "regions": {
          "note": "Die Quellenprüfung dieses Eintrags hat dieses Produktfeld nicht belegt."
        },
        "api": {
          "note": "Redaktionelle Einordnung dokumentierter Funktionen; kein Leistungstest."
        },
        "license": {
          "note": "Die Quellenprüfung dieses Eintrags hat dieses Produktfeld nicht belegt."
        },
        "commercial_use": {
          "note": "Die Quellenprüfung dieses Eintrags hat dieses Produktfeld nicht belegt."
        },
        "languages": {
          "note": "Die Quellenprüfung dieses Eintrags hat dieses Produktfeld nicht belegt."
        },
        "billing": {
          "note": "Redaktionelle Einordnung dokumentierter Funktionen; kein Leistungstest."
        }
      },
      "facts": {
        "bentoml.capability.primary": {
          "value": "Inferenz-APIs und mehrere Modelle kombinieren"
        }
      }
    },
    "pt": {
      "description": "Framework Python para empacotar modelos e inferência em serviços implantáveis.",
      "overview": "Usos comuns: Serviço de modelos para aplicações; Implantação de endpoints de inferência. Recursos documentados: APIs de inferência e composição de modelos; Dependências e modelos em contêineres reproduzíveis; Lotes adaptativos e filas assíncronas.",
      "features": [
        "APIs de inferência e composição de modelos",
        "Dependências e modelos em contêineres reproduzíveis",
        "Lotes adaptativos e filas assíncronas"
      ],
      "use_cases": [
        "Serviço de modelos para aplicações",
        "Implantação de endpoints de inferência"
      ],
      "pricing_notes": "Auto-hospedar o projeto aberto e contratar a plataforma hospedada são opções distintas. O serviço tem planos e limites próprios; a hospedagem própria exige infraestrutura e custos de modelos.",
      "faq": [
        {
          "q": "Em que parte do fluxo BentoML se encaixa?",
          "a": "Framework Python para empacotar modelos e inferência em serviços implantáveis. APIs de inferência e composição de modelos."
        },
        {
          "q": "Quais custos e limites comparar antes de escolher BentoML?",
          "a": "Auto-hospedar o projeto aberto e contratar a plataforma hospedada são opções distintas. O serviço tem planos e limites próprios; a hospedagem própria exige infraestrutura e custos de modelos."
        }
      ],
      "selection_evidence": {
        "tasks": {
          "note": "Classificação editorial dos recursos documentados; não é um teste de desempenho."
        },
        "form_factor": {
          "note": "Classificação editorial dos recursos documentados; não é um teste de desempenho."
        },
        "platforms": {
          "note": "A revisão de fontes desta entrada não estabeleceu este dado do produto."
        },
        "deployment": {
          "note": "Classificação editorial dos recursos documentados; não é um teste de desempenho."
        },
        "offline": {
          "note": "A revisão de fontes desta entrada não estabeleceu este dado do produto."
        },
        "regions": {
          "note": "A revisão de fontes desta entrada não estabeleceu este dado do produto."
        },
        "api": {
          "note": "Classificação editorial dos recursos documentados; não é um teste de desempenho."
        },
        "license": {
          "note": "A revisão de fontes desta entrada não estabeleceu este dado do produto."
        },
        "commercial_use": {
          "note": "A revisão de fontes desta entrada não estabeleceu este dado do produto."
        },
        "languages": {
          "note": "A revisão de fontes desta entrada não estabeleceu este dado do produto."
        },
        "billing": {
          "note": "Classificação editorial dos recursos documentados; não é um teste de desempenho."
        }
      },
      "facts": {
        "bentoml.capability.primary": {
          "value": "APIs de inferência e composição de modelos"
        }
      }
    },
    "ko": {
      "description": "AI 모델과 사용자 추론 파이프라인을 배포 서비스로 패키징하는 Python 프레임워크입니다.",
      "overview": "주요 용도는 앱에 AI 모델 제공; 운영 추론 엔드포인트 배포입니다. 문서에 명시된 기능: 추론 API 구축 및 다중 모델 구성; 종속성과 모델을 재현 가능 컨테이너로 패키징; 적응형 배치 및 비동기 작업 대기열.",
      "features": [
        "추론 API 구축 및 다중 모델 구성",
        "종속성과 모델을 재현 가능 컨테이너로 패키징",
        "적응형 배치 및 비동기 작업 대기열"
      ],
      "use_cases": [
        "앱에 AI 모델 제공",
        "운영 추론 엔드포인트 배포"
      ],
      "pricing_notes": "오픈 소스 자체 호스팅과 호스팅 플랫폼은 별도 선택지입니다. 호스팅 서비스에는 자체 요금제와 사용 제한이 있으며 자체 호스팅에도 인프라 및 모델 비용이 듭니다.",
      "faq": [
        {
          "q": "BentoML은(는) 워크플로의 어느 단계에 적합한가요?",
          "a": "AI 모델과 사용자 추론 파이프라인을 배포 서비스로 패키징하는 Python 프레임워크입니다. 추론 API 구축 및 다중 모델 구성."
        },
        {
          "q": "BentoML 선택 전에 어떤 비용과 제한을 비교해야 하나요?",
          "a": "오픈 소스 자체 호스팅과 호스팅 플랫폼은 별도 선택지입니다. 호스팅 서비스에는 자체 요금제와 사용 제한이 있으며 자체 호스팅에도 인프라 및 모델 비용이 듭니다."
        }
      ],
      "selection_evidence": {
        "tasks": {
          "note": "문서화된 기능에 따른 편집상 분류이며 성능 실측이 아닙니다."
        },
        "form_factor": {
          "note": "문서화된 기능에 따른 편집상 분류이며 성능 실측이 아닙니다."
        },
        "platforms": {
          "note": "이 항목의 출처 검토에서 해당 제품 정보를 확인하지 못했습니다."
        },
        "deployment": {
          "note": "문서화된 기능에 따른 편집상 분류이며 성능 실측이 아닙니다."
        },
        "offline": {
          "note": "이 항목의 출처 검토에서 해당 제품 정보를 확인하지 못했습니다."
        },
        "regions": {
          "note": "이 항목의 출처 검토에서 해당 제품 정보를 확인하지 못했습니다."
        },
        "api": {
          "note": "문서화된 기능에 따른 편집상 분류이며 성능 실측이 아닙니다."
        },
        "license": {
          "note": "이 항목의 출처 검토에서 해당 제품 정보를 확인하지 못했습니다."
        },
        "commercial_use": {
          "note": "이 항목의 출처 검토에서 해당 제품 정보를 확인하지 못했습니다."
        },
        "languages": {
          "note": "이 항목의 출처 검토에서 해당 제품 정보를 확인하지 못했습니다."
        },
        "billing": {
          "note": "문서화된 기능에 따른 편집상 분류이며 성능 실측이 아닙니다."
        }
      },
      "facts": {
        "bentoml.capability.primary": {
          "value": "추론 API 구축 및 다중 모델 구성"
        }
      }
    }
  },
  "verification": {
    "last_checked": "2026-11-03",
    "batch": "catalog-500-20261103",
    "source_review": "Official product pages read; no product-performance test.",
    "reviewed_sources": [
      "https://docs.bentoml.com/",
      "https://www.bentoml.com/"
    ],
    "egress_region": "unknown",
    "owner": "technical",
    "gap_note": "Only documented product scope and supplied claims were reviewed. Fields left unknown were not established for this entry; no launch date, geographic guarantee or licence is inferred."
  },
  "url": "https://globalaiproductindex.com/products/bentoml/"
}