Vertex AI 기반 OpenAI 모델은 완전 관리형 서버리스 모델을 API로 제공합니다. Vertex AI에서 OpenAI 모델을 사용하려면 요청을 Vertex AI API 엔드포인트로 직접 보냅니다. OpenAI 모델은 관리형 API를 사용하므로 인프라를 프로비저닝하거나 관리할 필요가 없습니다.
대답을 스트리밍하여 최종 사용자의 지연 시간 인식을 줄일 수 있습니다. 스트리밍된 응답은 서버 전송 이벤트 (SSE)를 사용하여 응답을 점진적으로 스트리밍합니다.
사용 가능한 OpenAI 모델
Vertex AI에 사용할 수 있는 OpenAI의 모델은 다음과 같습니다. OpenAI 모델에 액세스하려면 Model Garden 모델 카드로 이동합니다.
gpt-oss 120B
OpenAI gpt-oss 120B는 Apache 2.0 라이선스에 따라 출시된 1,200억 개의 가중치를 가진 언어 모델입니다. 추론 및 함수 호출 사용 사례에 적합합니다. 이 모델은 소비자 하드웨어에 배포하도록 최적화되어 있습니다.
120B 모델은 80GB GPU 하나에서 실행되면서 핵심 추론 벤치마크에서 OpenAI o4-mini와 거의 동등한 성능을 달성합니다.
Vertex AI에서 OpenAI 모델을 사용하려면 다음 단계를 실행해야 합니다. Vertex AI를 사용하려면 Vertex AI API(aiplatform.googleapis.com)를 사용 설정해야 합니다. Vertex AI API가 사용 설정된 기존 프로젝트가 이미 있는 경우 새 프로젝트를 만드는 대신 해당 프로젝트를 사용할 수 있습니다.
Sign in to your Google Cloud account. If you're new to
Google Cloud,
create an account to evaluate how our products perform in
real-world scenarios. New customers also get $300 in free credits to
run, test, and deploy workloads.
In the Google Cloud console, on the project selector page,
select or create a Google Cloud project.
[[["이해하기 쉬움","easyToUnderstand","thumb-up"],["문제가 해결됨","solvedMyProblem","thumb-up"],["기타","otherUp","thumb-up"]],[["이해하기 어려움","hardToUnderstand","thumb-down"],["잘못된 정보 또는 샘플 코드","incorrectInformationOrSampleCode","thumb-down"],["필요한 정보/샘플이 없음","missingTheInformationSamplesINeed","thumb-down"],["번역 문제","translationIssue","thumb-down"],["기타","otherDown","thumb-down"]],["최종 업데이트: 2025-09-04(UTC)"],[],[],null,["# OpenAI models\n\n| **Note:** OpenAI models are not a Google product, and its availability in Vertex AI is subject to the terms for \"Separate Offerings\" in the AI/ML Services section of the [Service Specific\n| Terms](/terms/service-terms), and separate terms found in the relevant model card.\n\nOpenAI models on Vertex AI offer fully managed and serverless\nmodels as APIs. To use an OpenAI model on Vertex AI, send\na request directly to the Vertex AI API endpoint. Because\nOpenAI models use a managed API, there's no need to provision or\nmanage infrastructure.\n\nYou can stream your responses to reduce the end-user latency perception. A\nstreamed response uses *server-sent events* (SSE) to incrementally stream the\nresponse.\n\nAvailable OpenAI models\n-----------------------\n\nThe following models are available from OpenAI to use in\nVertex AI. To access an OpenAI model, go to its\nModel Garden model card.\n\n### gpt-oss 120B\n\nOpenAI gpt-oss 120B is a 120B open-weight language model\nreleased under the Apache\n2.0 license. It is well-suited for reasoning and function calling use cases. The\nmodel is optimized for deployment on consumer hardware.\n\nThe 120B model achieves near-parity with OpenAI o4-mini on core reasoning\nbenchmarks, while running on a single 80GB GPU.\n\n[Go to the gpt-oss 120B model card](https://console.cloud.google.com/vertex-ai/publishers/openai/model-garden/gpt-oss-120b-maas)\n\n### gpt-oss 20B\n\nOpenAI gpt-oss 20B is a 20B open-weight language model\nreleased under the Apache\n2.0 license. It is well-suited for reasoning and function calling use cases. The\nmodel is optimized for deployment on consumer hardware.\n\nThe 20B model delivers similar results to OpenAI o3-mini on common benchmarks\nand can run on edge devices with 16GB of memory, making it ideal for on-device\nuse cases, local inference, or rapid iteration without costly infrastructure.\n\n[Go to the gpt-oss 20B model card](https://console.cloud.google.com/vertex-ai/publishers/openai/model-garden/gpt-oss-120b-maas)\n\n### Before you begin\n\nTo use OpenAI models with Vertex AI, you must perform the\nfollowing steps. The Vertex AI API\n(`aiplatform.googleapis.com`) must be enabled to use\nVertex AI. If you already have an existing project with the\nVertex AI API enabled, you can use that project instead of creating a\nnew project.\n\n- Sign in to your Google Cloud account. If you're new to Google Cloud, [create an account](https://console.cloud.google.com/freetrial) to evaluate how our products perform in real-world scenarios. New customers also get $300 in free credits to run, test, and deploy workloads.\n- In the Google Cloud console, on the project selector page,\n select or create a Google Cloud project.\n\n [Go to project selector](https://console.cloud.google.com/projectselector2/home/dashboard)\n-\n [Verify that billing is enabled for your Google Cloud project](/billing/docs/how-to/verify-billing-enabled#confirm_billing_is_enabled_on_a_project).\n\n-\n\n\n Enable the Vertex AI API.\n\n\n [Enable the API](https://console.cloud.google.com/flows/enableapi?apiid=aiplatform.googleapis.com)\n\n- In the Google Cloud console, on the project selector page,\n select or create a Google Cloud project.\n\n [Go to project selector](https://console.cloud.google.com/projectselector2/home/dashboard)\n-\n [Verify that billing is enabled for your Google Cloud project](/billing/docs/how-to/verify-billing-enabled#confirm_billing_is_enabled_on_a_project).\n\n-\n\n\n Enable the Vertex AI API.\n\n\n [Enable the API](https://console.cloud.google.com/flows/enableapi?apiid=aiplatform.googleapis.com)\n\n1. Go to one of the following Model Garden model cards, then click **Enable**."]]
