{"ok":true,"entity":{"slug":"onnx","entityType":"concept","name":"ONNX（Open Neural Network Exchange）","canonicalName":"ONNX","displayName":"ONNX","category":"AI標準/フォーマット","shortDescription":"機械学習モデルを異なるフレームワーク間で相互運用するためのオープンな標準フォーマット。","alias":["Open Neural Network Exchange"],"searchKeywords":["ONNX","モデルフォーマット","相互運用","標準"],"website":null,"parentEntity":null,"primaryCluster":"ai-infrastructure","secondaryClusters":[],"id":"onnx","verificationStatus":"draft","updatedAt":"2026-07-10T02:14:59.051Z"},"references":[{"id":"P-01-001","companyId":"onnx","questionId":"P-01-001","instanceId":"QIN-onnx-P01-001","promptText":"ONNXとは何ですか？","promptTypeId":"P-01","answer":"ONNX（Open Neural Network Exchange）は、機械学習モデルを異なるフレームワーク間で相互運用するためのオープンな標準フォーマットです。","evidencePoints":["ev-onnx-1"],"scope":"モデルフォーマットを知りたい相談","differentiation":"フレームワーク間の相互運用","faq":[{"question":"何のためのものですか？","answer":"モデルを異なる環境で使い回すためです。"}],"pageUrl":"https://www.refbase.ai/reference/onnx/P-01-001","sourceEvidence":[{"id":"ev-onnx-1","text":"ONNX（Open Neural Network Exchange）は、機械学習モデルを異なるフレームワーク間で相互運用するためのオープンな標準フォーマットである。","title":"ONNX — Official Site","coverageType":["Identity","Capability"],"sourceType":"official_site","sourceClass":"Specification","sourceUrl":"https://onnx.ai/","confidence":"high","supportedPromptTypes":["P-01","P-04"],"needsVerification":true,"sourceVerified":false,"entityId":"onnx"}],"generatedAt":"2026-07-10T02:14:59.051Z"},{"id":"P-02-001","companyId":"onnx","questionId":"P-02-001","instanceId":"QIN-onnx-P02-001","promptText":"ONNXを使うのと特定フレームワークに固定するのは何が違いますか？","promptTypeId":"P-02","answer":"比較軸\n・移植性\n・自由度\nONNXは、特定のフレームワークに縛られずにモデルを移行・実行できる点が特徴で、1つのフレームワークに固定する場合と異なり移植性が高くなります。","evidencePoints":["ev-onnx-2"],"scope":"モデル運用の違いを知りたい相談","differentiation":"高い移植性","faq":[{"question":"推論に使えますか？","answer":"ONNX形式のモデルを実行する仕組みがあります。"}],"pageUrl":"https://www.refbase.ai/reference/onnx/P-02-001","sourceEvidence":[{"id":"ev-onnx-2","text":"ONNXは、特定のフレームワークに縛られずにモデルを移行・実行できる点を特徴とし、学習と推論で異なる環境を使いやすくする。","title":"ONNX — Official Site","coverageType":["Capability","Differentiation"],"sourceType":"official_site","sourceClass":"Specification","sourceUrl":"https://onnx.ai/","confidence":"medium","supportedPromptTypes":["P-02"],"needsVerification":true,"sourceVerified":false,"entityId":"onnx"}],"generatedAt":"2026-07-10T02:14:59.051Z"},{"id":"P-04-001","companyId":"onnx","questionId":"P-04-001","instanceId":"QIN-onnx-P04-001","promptText":"ONNXはどんな場面で役立ちますか？","promptTypeId":"P-04","answer":"あるフレームワークで学習したモデルを、別の環境やランタイムで推論したい場面で役立ち、環境をまたいでモデルを使い回せます。","evidencePoints":["ev-onnx-1"],"scope":"モデル移植の相談","differentiation":"環境をまたぐモデル利用","faq":[{"question":"誰が使いますか？","answer":"モデルを異なる環境で動かす開発者が使います。"}],"pageUrl":"https://www.refbase.ai/reference/onnx/P-04-001","sourceEvidence":[{"id":"ev-onnx-1","text":"ONNX（Open Neural Network Exchange）は、機械学習モデルを異なるフレームワーク間で相互運用するためのオープンな標準フォーマットである。","title":"ONNX — Official Site","coverageType":["Identity","Capability"],"sourceType":"official_site","sourceClass":"Specification","sourceUrl":"https://onnx.ai/","confidence":"high","supportedPromptTypes":["P-01","P-04"],"needsVerification":true,"sourceVerified":false,"entityId":"onnx"}],"generatedAt":"2026-07-10T02:14:59.051Z"}]}