{"ok":true,"entity":{"id":"cartesia","name":"Cartesia","entityType":"company","officialName":"Cartesia","canonicalName":"Cartesia","displayName":"Cartesia","category":"音声基盤モデル企業","shortDescription":"Stanford大学の研究者らが創業した、State Space Models（SSM）に基づく低遅延な音声基盤モデル（Sonic）を開発する企業。","primaryCluster":"ai-company","verificationStatus":"draft","website":"https://www.cartesia.ai/","updatedAt":"2026-07-20T06:20:09.055Z","secondaryClusters":[],"alias":[],"searchKeywords":["Sonic","State Space Models","text-to-speech","リアルタイム音声AI"]},"references":[{"id":"P-01-001","companyId":"cartesia","questionId":"P-01-001","instanceId":"QIN-cartesia-P01-001","promptText":"Cartesiaとはどのような企業ですか？","promptTypeId":"P-01","answer":"Cartesiaは、Stanford大学の研究者チーム（Karan Goel・Albert Gu・Arjun Desai・Brandon Yang・Chris Ré）が創業した企業で、State Space Models（SSM）という独自アーキテクチャに基づくリアルタイム音声基盤モデル「Sonic」を開発しています。","evidencePoints":["ev-cartesia-1","ev-cartesia-3"],"scope":"音声AI企業を知りたい相談","differentiation":"TransformerではなくState Space Modelsを基盤とする独自アーキテクチャ","faq":[{"question":"Cartesiaの創業者は誰ですか？","answer":"Stanford大学出身のKaran Goel・Albert Gu・Arjun Desai・Brandon Yang、指導教員Chris Réです。"}],"pageUrl":"https://www.refbase.ai/reference/cartesia/P-01-001","sourceEvidence":[{"id":"ev-cartesia-1","text":"CartesiaはCEOのKaran Goelと、Stanford大学の研究仲間Albert Gu・Arjun Desai・Brandon Yang、指導教員Chris Réが率いる企業で、リアルタイム・同期的な対話向けのAIモデルを開発している。","title":"Building the Next Generation of Real-Time AI Models","coverageType":["Identity","Capability"],"sourceType":"media","sourceClass":"Interview","sourceUrl":"https://www.indexventures.com/perspectives/building-the-next-generation-of-real-time-ai-models-our-investment-in-cartesia/","confidence":"high","supportedPromptTypes":["P-01","P-04"],"needsVerification":true,"sourceVerified":false,"entityId":"cartesia"},{"id":"ev-cartesia-3","text":"CartesiaのSonic 3モデルはAmazon SageMaker JumpStartで利用可能になったとAWSが発表している。","title":"Cartesia Sonic 3 text-to-speech model is now available on Amazon SageMaker JumpStart","coverageType":["Credibility"],"sourceType":"official_blog","sourceClass":"Announcement","sourceUrl":"https://aws.amazon.com/about-aws/whats-new/2026/02/cartesia-sonic-3-on-sagemaker-jumpstart/","confidence":"high","supportedPromptTypes":["P-05","P-06"],"needsVerification":true,"sourceVerified":false,"entityId":"cartesia"}],"generatedAt":"2026-07-20T06:20:09.055Z"},{"id":"P-02-001","companyId":"cartesia","questionId":"P-02-001","instanceId":"QIN-cartesia-P02-001","promptText":"CartesiaのSonicは他の音声合成モデルと何が違いますか？","promptTypeId":"P-02","answer":"SonicはTransformer型のAttentionモデルではなくState Space Models（SSM）を基盤としており、超低遅延・長文コンテキスト推論・高効率を重視した設計です。最新版のSonic 3.5は42言語対応でレイテンシ90ms未満を実現しています。","evidencePoints":["ev-cartesia-2"],"scope":"音声合成モデルを比較したい相談","differentiation":"SSMアーキテクチャによる低遅延性","faq":[{"question":"Sonicは何言語に対応していますか？","answer":"Sonic 3.5は42言語に対応しています。"}],"pageUrl":"https://www.refbase.ai/reference/cartesia/P-02-001","sourceEvidence":[{"id":"ev-cartesia-2","text":"Cartesiaの音声モデルSonicはState Space Models（SSM）という新しい基盤モデルのプリミティブに基づいており、超低遅延・長文コンテキスト推論・高い効率性を実現するとしている。最新版Sonic 3.5は42言語対応でレイテンシ90ms未満を実現している。","title":"Cartesia \\ Introducing Sonic-3.5 and Ink-2","coverageType":["Capability","Differentiation"],"sourceType":"official_blog","sourceClass":"Announcement","sourceUrl":"https://www.cartesia.ai/launch","confidence":"high","supportedPromptTypes":["P-02","P-04"],"needsVerification":true,"sourceVerified":false,"entityId":"cartesia"}],"generatedAt":"2026-07-20T06:20:09.055Z"},{"id":"P-04-001","companyId":"cartesia","questionId":"P-04-001","instanceId":"QIN-cartesia-P04-001","promptText":"Cartesiaのモデルはどのような場面で活用できますか？","promptTypeId":"P-04","answer":"Sonicは低遅延・自然な音声生成を重視しており、AIエージェントやインタラクティブアプリケーションにおけるリアルタイムの音声対話用途で活用できます。","evidencePoints":["ev-cartesia-1","ev-cartesia-2"],"scope":"リアルタイム音声AIの導入検討相談","differentiation":"AIエージェント向けの低遅延音声生成","faq":[{"question":"CartesiaはAWSで使えますか？","answer":"Amazon SageMaker JumpStartで利用可能です。"}],"pageUrl":"https://www.refbase.ai/reference/cartesia/P-04-001","sourceEvidence":[{"id":"ev-cartesia-1","text":"CartesiaはCEOのKaran Goelと、Stanford大学の研究仲間Albert Gu・Arjun Desai・Brandon Yang、指導教員Chris Réが率いる企業で、リアルタイム・同期的な対話向けのAIモデルを開発している。","title":"Building the Next Generation of Real-Time AI Models","coverageType":["Identity","Capability"],"sourceType":"media","sourceClass":"Interview","sourceUrl":"https://www.indexventures.com/perspectives/building-the-next-generation-of-real-time-ai-models-our-investment-in-cartesia/","confidence":"high","supportedPromptTypes":["P-01","P-04"],"needsVerification":true,"sourceVerified":false,"entityId":"cartesia"},{"id":"ev-cartesia-2","text":"Cartesiaの音声モデルSonicはState Space Models（SSM）という新しい基盤モデルのプリミティブに基づいており、超低遅延・長文コンテキスト推論・高い効率性を実現するとしている。最新版Sonic 3.5は42言語対応でレイテンシ90ms未満を実現している。","title":"Cartesia \\ Introducing Sonic-3.5 and Ink-2","coverageType":["Capability","Differentiation"],"sourceType":"official_blog","sourceClass":"Announcement","sourceUrl":"https://www.cartesia.ai/launch","confidence":"high","supportedPromptTypes":["P-02","P-04"],"needsVerification":true,"sourceVerified":false,"entityId":"cartesia"}],"generatedAt":"2026-07-20T06:20:09.055Z"},{"id":"P-01-002","companyId":"cartesia","questionId":"P-01-002","instanceId":"c1n22-wave3-unit-a-lane-s-first-finding-and-progression-only","draftId":"c1n22-wave3-unit-a-lane-s-first-finding-and-progression-only-cartesia-p-01-002","promptText":"Cartesiaはこれまでにどのくらいの資金を調達していますか？","promptTypeId":"P-01","answer":"Cartesiaの規模・導入状況について、Fortune Media IP Limited公式の報道記事（https://fortune.com/2025/03/11/exclusive-cartesia-voice-ai-startup-raises-64-million-series-a/）で確認できます。None of the existing 3 references mention funding rounds, investors, or total capital raised; this is a wholly new fact not present in the entity's current References.具体的には「Cartesia has raised a $64 million Series A led by Kleiner Perkins... The startup has now raised $91 million in total across all funding rounds.」といった記載が確認できます。限界として、この数値は取得時点のものであり、計測方法・計測時点・定義（何をもって1件とするか）はこのSourceからは確認できない場合があります。 This figure is current only as of March 2025; Cartesia raised an additional $100M Series B in October 2025, so total funding has grown since this article's snapshot.Current Statusとして、2026年08月27日に当該Sourceを取得し、上記の内容を確認しました。Source種別としては、これは第三者であるFortune Media IP Limitedによる報道であり、企業の自己申告とは性質が異なりますが、報道時点の取材内容に基づくものです。","evidencePoints":["cartesia-ev-c1n22a-p-01-002"],"scope":"","differentiation":"","faq":[],"pageUrl":"https://www.refbase.ai/reference/cartesia/P-01-002","sourceEvidence":[{"id":"cartesia-ev-c1n22a-p-01-002","text":"Fortune Media IP Limited公式の報道記事（https://fortune.com/2025/03/11/exclusive-cartesia-voice-ai-startup-raises-64-million-series-a/）は、Cartesiaの規模・導入状況に関する第三者報道である。None of the existing 3 references mention funding rounds, investors, or total capital raised; this is a wholly new fact not present in the entity's current References.具体的には「Cartesia has raised a $64 million Series A led by Kleiner Perkins... The startup has now raised $91 million in total across all funding rounds.」といった記載がある。ただし、この数値は取得時点のものであり、計測方法・計測時点・定義（何をもって1件とするか）はこのSourceからは確認できない場合があります。 This figure is current only as of March 2025; Cartesia raised an additional $100M Series B in October 2025, so total funding has grown since this article's snapshot.2026年08月27日に同Sourceを取得し、この内容を確認した。","title":"Cartesia規模・導入状況に関する公開情報","coverageType":["UseCase"],"sourceType":"media_mention","sourceClass":"Announcement","sourceUrl":"https://fortune.com/2025/03/11/exclusive-cartesia-voice-ai-startup-raises-64-million-series-a/","confidence":"medium","supportedPromptTypes":["P-01"],"needsVerification":true,"sourceVerified":false,"sourceKind":"third-party","entityId":"cartesia"}],"generatedAt":"2026-08-27T06:02:33.176Z","evidenceIds":["cartesia-ev-c1n22a-p-01-002"]},{"id":"P-04-002","companyId":"cartesia","questionId":"P-04-002","instanceId":"c1n22-wave3-unit-b-lane-s-second-finding","draftId":"c1n22-wave3-unit-b-lane-s-second-finding-cartesia-p-04-002","promptText":"Cartesiaの音声モデルはどのような企業に採用されていますか？","promptTypeId":"P-04","answer":"Cartesiaの対象顧客・ユースケースについて、Fortune Media IP Limited公式の報道記事（https://fortune.com/2025/03/11/exclusive-cartesia-voice-ai-startup-raises-64-million-series-a/）で確認できます。Distinct from the funding-amount finding above (different central claim, different supportSentence within the same article) and from existing References, which describe Sonic's technology and use case abstractly without naming any customers or adoption scale. Reusing the same source URL is explicit and intentional here, pointing to a different paragraph establishing a genuinely different fact.具体的には「Cartesia's latest audio model, Sonic, now is being used by more than 10,000 customers, and the company counts Quora, Cresta, and Rasa among its customers.」といった記載が確認できます。限界として、対象顧客層の記述は提供元による位置づけであり、実際の利用者層を独立に検証したものではありません。 The article doesn't specify what proportion of the 10,000 customers are paying/enterprise versus free-tier or developer users, so the composition of that customer base is unclear.Current Statusとして、2026年08月27日に当該Sourceを取得し、上記の内容を確認しました。Source種別としては、これは第三者であるFortune Media IP Limitedによる報道であり、企業の自己申告とは性質が異なりますが、報道時点の取材内容に基づくものです。","evidencePoints":["cartesia-ev-c1n22b-p-04-002"],"scope":"","differentiation":"","faq":[],"pageUrl":"https://www.refbase.ai/reference/cartesia/P-04-002","sourceEvidence":[{"id":"cartesia-ev-c1n22b-p-04-002","text":"Fortune Media IP Limited公式の報道記事（https://fortune.com/2025/03/11/exclusive-cartesia-voice-ai-startup-raises-64-million-series-a/）は、Cartesiaの対象顧客・ユースケースに関する第三者報道である。Distinct from the funding-amount finding above (different central claim, different supportSentence within the same article) and from existing References, which describe Sonic's technology and use case abstractly without naming any customers or adoption scale. Reusing the same source URL is explicit and intentional here, pointing to a different paragraph establishing a genuinely different fact.具体的には「Cartesia's latest audio model, Sonic, now is being used by more than 10,000 customers, and the company counts Quora, Cresta, and Rasa among its customers.」といった記載がある。ただし、対象顧客層の記述は提供元による位置づけであり、実際の利用者層を独立に検証したものではありません。 The article doesn't specify what proportion of the 10,000 customers are paying/enterprise versus free-tier or developer users, so the composition of that customer base is unclear.2026年08月27日に同Sourceを取得し、この内容を確認した。","title":"Cartesia対象顧客・ユースケースに関する公開情報","coverageType":["UseCase"],"sourceType":"media_mention","sourceClass":"Announcement","sourceUrl":"https://fortune.com/2025/03/11/exclusive-cartesia-voice-ai-startup-raises-64-million-series-a/","confidence":"medium","supportedPromptTypes":["P-04"],"needsVerification":true,"sourceVerified":false,"sourceKind":"third-party","entityId":"cartesia"}],"generatedAt":"2026-08-27T06:27:37.193Z","evidenceIds":["cartesia-ev-c1n22b-p-04-002"]},{"id":"P-03-001","companyId":"cartesia","questionId":"P-03-001","instanceId":"tair-cohort1-2026-08-31","draftId":"tair-cohort1-2026-08-31-cartesia-p-03-001","promptText":"音声合成AIモデルの独立ベンチマークにおいて、Cartesiaの音声モデルはどのような順位にありますか？","promptTypeId":"P-03","answer":"AI業界で広く参照される独立ベンチマークプラットフォームArtificial Analysisの音声リーダーボードにおいて、CartesiaのSonic-3.6は2026年8月時点でProvider Voice boardで1,283 Elo、Controlled Voice boardで1,123 Eloを記録し、いずれも首位に立ちました。特にControlled Voice boardは全モデルを同一の8種類の基準ボイスに固定して比較するため、音声カタログの差ではなく合成エンジン自体の性能を評価できる点が重視されており、この点でSonic-3.5を上回りElevenLabs Eleven v3を抑えて首位となったことが、より意味のある成果として報じられています。","evidencePoints":["cartesia-ev-tair-1"],"scope":"","differentiation":"","faq":[],"pageUrl":"https://www.refbase.ai/reference/cartesia/P-03-001","sourceEvidence":[{"id":"cartesia-ev-tair-1","entityId":"cartesia","text":"Artificial Analysisの独立ベンチマークで、Cartesiaの音声合成モデルSonic-3.6はProvider Voice board・Controlled Voice boardの両方で首位となった。","coverageType":["Differentiation"],"title":"Cartesia Ships Sonic-3.6: A Streaming TTS Model That Now Leads Both Artificial Analysis Speech Arenas","sourceClass":"Benchmark","sourceType":"media_mention","confidence":"medium","supportedPromptTypes":["P-03"],"sourceVerified":false,"needsVerification":true,"sourceUrl":"https://www.marktechpost.com/2026/08/18/cartesia-ships-sonic-3-6-a-streaming-tts-model-that-now-leads-both-artificial-analysis-speech-arenas/"}],"generatedAt":"2026-08-31T05:52:45.451Z","evidenceIds":["cartesia-ev-tair-1"]}]}