콘텐츠로 건너뛰기
  • 모델
  • 순위
  • Ori
ElevenLabs launch offer: every ElevenLabs model is 50% off through October 19, 2026. See ElevenLabs models

Footer

OpenRouterOpenRouter
© 2026 OpenRouter, Inc

제품

  • 채팅
  • 순위
  • 벤치마크
  • 앱
  • 찾아보기
  • 모델
  • Ori
  • 컬렉션
  • 제공업체
  • 도구
  • 가격
  • Business
  • Enterprise
  • 랩

회사

  • 정보
  • 블로그
  • 채용
    채용
  • 개인정보 보호
  • 서비스 약관
  • 신뢰 센터
  • 지원
  • OR과 호환
  • 데이터
  • 브랜드

개발자

  • 문서
  • API 참조
  • 개발자 플랫폼
  • 상태
  • AI 사이트맵

연결

  • Discord
  • GitHub
  • LinkedIn
  • X
  • YouTube
openai의 파비콘

OpenAI: gpt-oss-20b

openai/gpt-oss-20b

모델 가중치
비교

gpt-oss-20b is an open-weight 21B parameter model released by OpenAI under the Apache 2.0 license. It uses a Mixture-of-Experts (MoE) architecture with 3.6B active parameters per forward pass, optimized for lower-latency inference and deployability on consumer or single-GPU hardware. The model is trained in OpenAI’s Harmony response format and supports reasoning level configuration, fine-tuning, and agentic capabilities including function calling, tool use, and structured outputs.

모달리티
입력/출력 가격
US$0.018 / US$0.091M당
컨텍스트
131K
출시
2025년 8월 5일
지식 기준일
2024년 6월
비교
제공업체가격성능가동 시간벤치마크앱활동자주 묻는 질문탐색

제공업체

같은 모델을 서로 다른 회사가 호스팅합니다. OpenRouter는 사용자가 선택한 라우팅 모드에 따라 요청을 그중 한 곳으로 라우팅합니다 — Balanced(가격 + 속도), Nitro(가장 빠름), Floor(가장 저렴함) 또는 Exacto(가장 높은 도구 호출 정확도).

가격

고객이 이 모델에 실제로 지불하는 평균 가격으로, 제공업체가 게시한 가격과 함께 표시됩니다. 캐싱과 할인으로 인해 실제 지불 가격은 게시된 가격보다 훨씬 낮은 경우가 많습니다.

성능

처리량은 모델이 얼마나 빠르게 출력하는지를 나타냅니다(초당 토큰 수 — 높을수록 좋음). 지연 시간은 전체 왕복 시간입니다(낮을수록 좋음). TTFT는 첫 토큰까지 걸리는 시간으로, 무언가 화면에 표시되기 시작할 때까지의 시간을 뜻합니다(낮을수록 좋음).

가동 시간

가동 시간은 지난 3일 동안 하나 이상의 제공업체가 요청에 응답한 시간의 비율입니다. 가용성은 추론이 성공적으로 제공된 시간의 비율입니다. OpenRouter는 지속적으로 모니터링하며, 한 제공업체에서 오류를 반환하면 다음으로 적합한 제공업체를 사용합니다.

벤치마크

표준화된 평가와 공유 벤치마크 프롬프트에 대한 이 모델 자체의 출력 점수입니다. 백분율이 높을수록 더 좋으며, 각 순위는 이 모델이 해당 출처에서 평가된 모델 중 어디에 위치하는지를 보여줍니다.

벤치마크 점수 요약: OpenAI: gpt-oss-20b (Artificial Analysis, Design Arena and Vals AI)
출처벤치마크점수
Artificial Analysisgpt-oss-20b (Low) Intelligence Index10.0
Artificial Analysisgpt-oss-20b (Low) Math Index62.3
Artificial Analysisgpt-oss-20b (Low) GPQA Diamond61.1%
Artificial Analysisgpt-oss-20b (Low) HLE5.3%
Artificial Analysisgpt-oss-20b (Low) IFBench57.8%
Artificial Analysisgpt-oss-20b (Low) τ²-Bench Telecom50.3%
Artificial Analysisgpt-oss-20b (Low) AA-LCR31.0%
Artificial Analysisgpt-oss-20b (Low) CritPt0.0%
Artificial Analysisgpt-oss-20b (Low) LiveCodeBench65.2%
Artificial Analysisgpt-oss-20b (Low) Terminal-Bench Hard4.5%
Artificial Analysisgpt-oss-20b (Low) MMLU-Pro71.8%
Artificial Analysisgpt-oss-20b (Low) AA-Omniscience Accuracy15.1%
Artificial Analysisgpt-oss-20b (Low) AA-Omniscience Non-Hallucination Rate13.3%
Artificial Analysisgpt-oss-20b (Low) AIME 202562.3%
Artificial Analysisgpt-oss-20b (High) Intelligence Index9.0
Artificial Analysisgpt-oss-20b (High) Coding Index20.7
Artificial Analysisgpt-oss-20b (High) Math Index89.3
Artificial Analysisgpt-oss-20b (High) Agentic Index1.2
Artificial Analysisgpt-oss-20b (High) GPQA Diamond68.8%
Artificial Analysisgpt-oss-20b (High) HLE11.0%
Artificial Analysisgpt-oss-20b (High) IFBench65.1%
Artificial Analysisgpt-oss-20b (High) τ²-Bench Telecom60.2%
Artificial Analysisgpt-oss-20b (High) AA-LCR34.7%
Artificial Analysisgpt-oss-20b (High) τ-Bench Banking7.0%
Artificial Analysisgpt-oss-20b (High) GDPval-AA0.0%
Artificial Analysisgpt-oss-20b (High) CritPt1.4%
Artificial Analysisgpt-oss-20b (High) LiveCodeBench77.7%
Artificial Analysisgpt-oss-20b (High) SciCode38.9%
Artificial Analysisgpt-oss-20b (High) Terminal-Bench Hard10.6%
Artificial Analysisgpt-oss-20b (High) Terminal-Bench 4.00.0%
Artificial Analysisgpt-oss-20b (High) Terminal-Bench 2.113.9%
Artificial Analysisgpt-oss-20b (High) MMLU-Pro74.8%
Artificial Analysisgpt-oss-20b (High) AA-Omniscience Accuracy16.0%
Artificial Analysisgpt-oss-20b (High) AA-Omniscience Non-Hallucination Rate5.9%
Artificial Analysisgpt-oss-20b (High) AIME 202589.3%
Design ArenaGPT OSS 20B Models Arena Data Visualization Elo935
Design ArenaGPT OSS 20B Models Arena Website Elo857
Vals AIGPT OSS 20B AIME86.0%
Vals AIGPT OSS 20B CaseLaw v243.8%
Vals AIGPT OSS 20B CorpFin v253.1%
Vals AIGPT OSS 20B GPQA Diamond68.9%
Vals AIGPT OSS 20B LiveCodeBench80.4%
Vals AIGPT OSS 20B LegalBench70.8%
Vals AIGPT OSS 20B MATH 50094.2%
Vals AIGPT OSS 20B MedQA82.9%
Vals AIGPT OSS 20B MGSM89.0%
Vals AIGPT OSS 20B MMLU Pro71.6%
Vals AIGPT OSS 20B TaxEval v263.7%

앱

이 모델로 가장 많은 트래픽을 보내는 공개 앱입니다. 실제 프로덕션 워크로드가 어떤 모습인지 파악하는 데 유용하며, 이 모델이 어떤 사용 사례에 가장 적합한지 알려 주는 단서이기도 합니다.

활동

시간 경과에 따른 이 모델의 토큰 사용량과 요청 트래픽입니다.

API

이 모델을 호출하는 데 바로 사용할 수 있는 코드입니다. OpenRouter의 API는 OpenAI와 호환되므로 대부분의 SDK는 기본 URL만 바꾸면 작동합니다. 모델 간에 변경되는 것은 아래의 모델 슬러그뿐입니다.

더 많은 모델 살펴보기

AI 모델 순위순위
US$0.018US$0.09US$0.00902.53s21 tps
100.00%
US$0.020US$0.10--0.75s29 tps
99.73%
US$0.029US$0.14US$0.02900.62s22 tps
99.98%
US$0.030US$0.13US$0.03000.22s148 tps
99.81%
US$0.030US$0.14--0.54s71 tps
99.48%
US$0.030US$0.15US$0.02002.51s43 tps
99.83%
US$0.070US$0.15--0.48s362 tps
99.78%
US$0.075US$0.30US$0.03750.36s432 tps
67.49%
US$0.040US$0.18--1.11s29 tps
87.00%
US$0.070US$0.25--0.44s176 tps
94.99%
Standard 라우팅에서 사용되지 않음:이 엔드포인트를 사용하지 않는 이유
US$0.070US$0.15--0.51s148 tps
100.00%
처리량
432tok/s

P50, 제공업체 전체 중 최고

지연 시간
0.22초

P50, 최고 제공업체

AutoExacto benchmarks
GPQA DiamondTAU-BenchDekaLLM69.5%--DeepInfra68.7%--AkashML67.8%--SiliconFlow63.7%--auto-routing65.2%55.0%
+8 개 추가 제공업체
가동 시간(3일)모델에 연결할 수 있습니다. 요청이 제공업체로 라우팅되었습니다.
100.00%
가용성(3일)모델이 어떤 제공업체에서든 추론 결과를 반환한 비율입니다. 오류와 빈 응답은 가동 시간을 낮춥니다.
96.20%

지난 3일간의 가용성

최근 72시간
가용성 96.20%
3일 전2일 전어제지금

지난 24시간 동안의 가용성

OpenRouter Availability
98.02%
Without Routing
90.99%

업스트림 제공업체에서 오류가 발생하면 요청 필터에서 허용하는 경우 정상 상태인 다른 제공업체로 라우팅하여 복구할 수 있습니다. 제공업체별 가동 시간 데이터는 프로그래밍 방식으로 엔드포인트 API를 통해 확인할 수 있습니다. 로드 밸런싱 및 사용자 지정 옵션에 대해 자세히 알아보세요.

1.
https://heyditto.ai/의 파비콘
HeyDitto
신규
149B토큰
2.
https://jobleads.com/의 파비콘
JobLeads LLM
Use AI to find jobs for you
32.9B토큰
3.
https://centralcommand.io/의 파비콘
Central Command
신규
28.5B토큰
4.
https://craftrpgs.com/?utm_source=openrouter의 파비콘
Craft
Build and play AI RPGs
14.1B토큰
5.
https://fvchat.app/의 파비콘
FVChat
신규
12.4B토큰

자주 묻는 질문

gpt-oss-20b is an open-weight 21B parameter model released by OpenAI under the Apache 2.0 license. It uses a Mixture-of-Experts (MoE) architecture with 3.6B active parameters per forward pass, optimized for lower-latency inference and deployability on consumer or single-GPU hardware.

gpt-oss-20b의 비용은 US$0.018/M 입력 토큰 및 US$0.09/M 출력 토큰, US$0.009/M tokens의 Cache Read에 대한 요금은 별도입니다.

gpt-oss-20b의 컨텍스트 윈도우는 131,072토큰입니다. 최대 32,768개의 완료 토큰을 지원합니다.

예. gpt-oss-20b은(는) 해당 모델을 제공하는 10개 제공업체 중 8곳에서 함수 호출을 위한 tools 및 tool_choice를 지원하며, 도구를 전송하는 요청은 해당 제공업체로 라우팅됩니다. 또한 response_format에서 JSON 스키마를 사용하여 구조화된 출력을 지원합니다.

gpt-oss-20b은(는) OpenRouter에서 10개 제공업체가 제공합니다: Darkbloom, AkashML, DekaLLM, CoreWeave, DeepInfra, Parasail, SiliconFlow, Amazon Bedrock 외 2개. 요청은 사용 가능한 최적의 제공업체로 라우팅되며, 다른 제공업체로 자동 장애 조치됩니다. 제공업체 라우팅을 사용하여 제공업체를 고정하거나 제외할 수도 있습니다.

gpt-oss-20b은(는) 2025년 8월 5일에 출시되었습니다. 지식 기준일은 2024년 6월 30일입니다.

다음 제공업체의 추가 모델 OpenAI

GPT-6 Luna Decisions

GPT-6 Luna Decisions is GPT-6 Luna served through OpenAI's Decisions API. Instead of generating text, it reads the content passed as state (text, JSON, or images) and returns typed, probabilistic answers to named questions in a single request: the probability of yes for a yes/no question (noul), a probability for every option plus the most likely one (choice), or a probability for every level of an ordered rubric plus the expected score (score).

It is built for classification, routing, moderation, and rubric grading where application code thresholds the returned numbers rather than parsing a chat reply. A request can carry up to 200 questions about the same content.

결정US$0.10 / US$0
GPT-6.1 Sol Pro

GPT-6.1 Sol Pro is the same underlying model as GPT-6.1 Sol, served with reasoning.mode set to pro for higher-quality responses on complex tasks.

Cost note: pro mode spends far more reasoning tokens per request, so a typical request costs several times more than the same request on GPT-6.1 Sol and takes much longer to complete. It is intended for hard, high-stakes problems where the extra accuracy justifies the cost. For everyday coding, agentic, and chat workloads, use GPT-6.1 Sol instead.

Learn more in OpenAI's docs: https://developers.openai.com/api/docs/guides/reasoning#reasoning-mode

텍스트컨텍스트 1.1MUS$2 / US$10
GPT-6.1 Sol Pro

GPT-6.1 Sol Pro is the same underlying model as GPT-6.1 Sol, served with reasoning.mode set to pro for higher-quality responses on complex tasks.

Cost note: pro mode spends far more reasoning tokens per request, so a typical request costs several times more than the same request on GPT-6.1 Sol and takes much longer to complete. It is intended for hard, high-stakes problems where the extra accuracy justifies the cost. For everyday coding, agentic, and chat workloads, use GPT-6.1 Sol instead.

Learn more in OpenAI's docs: https://developers.openai.com/api/docs/guides/reasoning#reasoning-mode

텍스트컨텍스트 1.1MUS$1 / US$5
GPT-6.1 Sol

GPT-6.1 Sol is an upgrade to GPT-6 Sol from OpenAI, positioned below the flagship GPT-6 Astra in the GPT-6 series. It is suited for agentic coding, computer use, document-heavy professional work, and multi-step business workflow automation, and approaches Astra-level results on these tasks at a much lower cost. Compared with GPT-6 Sol, it makes fewer factual errors and is more reliable at respecting explicit restrictions and user intent during agentic tasks.

텍스트컨텍스트 1.1MUS$2 / US$10
GPT-6.1 Sol

GPT-6.1 Sol is an upgrade to GPT-6 Sol from OpenAI, positioned below the flagship GPT-6 Astra in the GPT-6 series. It is suited for agentic coding, computer use, document-heavy professional work, and multi-step business workflow automation, and approaches Astra-level results on these tasks at a much lower cost. Compared with GPT-6 Sol, it makes fewer factual errors and is more reliable at respecting explicit restrictions and user intent during agentic tasks.

텍스트컨텍스트 1.1MUS$1 / US$5
GPT-6 Luna Pro

GPT-6 Luna Pro is the same underlying model as GPT-6 Luna, served with reasoning.mode set to pro for higher-quality responses on complex tasks.

Cost note: pro mode spends far more reasoning tokens per request, so a typical request costs several times more than the same request on GPT-6 Luna and takes much longer to complete. It is intended for hard, high-stakes problems where the extra accuracy justifies the cost. For everyday coding, agentic, and chat workloads, use GPT-6 Luna instead.

Learn more in OpenAI's docs: https://developers.openai.com/api/docs/guides/reasoning#reasoning-mode

텍스트컨텍스트 1.1MUS$0.10 / US$0.50
GPT-6 Luna Pro

GPT-6 Luna Pro is the same underlying model as GPT-6 Luna, served with reasoning.mode set to pro for higher-quality responses on complex tasks.

Cost note: pro mode spends far more reasoning tokens per request, so a typical request costs several times more than the same request on GPT-6 Luna and takes much longer to complete. It is intended for hard, high-stakes problems where the extra accuracy justifies the cost. For everyday coding, agentic, and chat workloads, use GPT-6 Luna instead.

Learn more in OpenAI's docs: https://developers.openai.com/api/docs/guides/reasoning#reasoning-mode

텍스트컨텍스트 1.1MUS$0.05 / US$0.25
GPT-6 Luna

GPT-6 Luna is the fast, cost-efficient model in OpenAI's GPT-6 series, positioned below GPT-6 Sol. It is suited for high-volume and latency-sensitive workloads such as chat, classification, and lightweight agentic tasks, and at higher reasoning effort it can take on complex software engineering and computer-use tasks that previously called for a Sol-tier model. It shares the GPT-6 family's gains in factual reliability and its clearer, more concise communication style.

텍스트컨텍스트 1.1MUS$0.10 / US$0.50
GPT-6 Luna

GPT-6 Luna is the fast, cost-efficient model in OpenAI's GPT-6 series, positioned below GPT-6 Sol. It is suited for high-volume and latency-sensitive workloads such as chat, classification, and lightweight agentic tasks, and at higher reasoning effort it can take on complex software engineering and computer-use tasks that previously called for a Sol-tier model. It shares the GPT-6 family's gains in factual reliability and its clearer, more concise communication style.

텍스트컨텍스트 1.1MUS$0.05 / US$0.25
GPT-6 Sol Pro

GPT-6 Sol Pro is the same underlying model as GPT-6 Sol, served with reasoning.mode set to pro for higher-quality responses on complex tasks.

Cost note: pro mode spends far more reasoning tokens per request, so a typical request costs several times more than the same request on GPT-6 Sol and takes much longer to complete. It is intended for hard, high-stakes problems where the extra accuracy justifies the cost. For everyday coding, agentic, and chat workloads, use GPT-6 Sol instead.

Learn more in OpenAI's docs: https://developers.openai.com/api/docs/guides/reasoning#reasoning-mode

텍스트컨텍스트 1.1MUS$2 / US$10
GPT-6 Sol Pro

GPT-6 Sol Pro is the same underlying model as GPT-6 Sol, served with reasoning.mode set to pro for higher-quality responses on complex tasks.

Cost note: pro mode spends far more reasoning tokens per request, so a typical request costs several times more than the same request on GPT-6 Sol and takes much longer to complete. It is intended for hard, high-stakes problems where the extra accuracy justifies the cost. For everyday coding, agentic, and chat workloads, use GPT-6 Sol instead.

Learn more in OpenAI's docs: https://developers.openai.com/api/docs/guides/reasoning#reasoning-mode

텍스트컨텍스트 1.1MUS$1 / US$5
GPT-6 Sol

GPT-6 Sol is the cost-efficient high-end model in OpenAI's GPT-6 series, positioned below the flagship GPT-6 Astra and above the fast GPT-6 Luna tier. It is suited for demanding professional work, agentic coding, business workflow automation, and computer use, and is particularly strong at long-horizon software engineering tasks in real codebases. It approaches Astra-level factual reliability at a much lower cost and shares Astra's clearer, more concise communication style in technical and coding conversations.

텍스트컨텍스트 1.1MUS$2 / US$10
GPT-6 Sol

GPT-6 Sol is the cost-efficient high-end model in OpenAI's GPT-6 series, positioned below the flagship GPT-6 Astra and above the fast GPT-6 Luna tier. It is suited for demanding professional work, agentic coding, business workflow automation, and computer use, and is particularly strong at long-horizon software engineering tasks in real codebases. It approaches Astra-level factual reliability at a much lower cost and shares Astra's clearer, more concise communication style in technical and coding conversations.

텍스트컨텍스트 1.1MUS$1 / US$5
GPT Astra Latest

This model always redirects to the latest model in the GPT Astra family.

텍스트컨텍스트 1.1M
GPT Sol Latest

This model always redirects to the latest model in the GPT Sol family.

텍스트컨텍스트 1.1M
GPT Terra Latest

This model always redirects to the latest model in the GPT Terra family.

텍스트컨텍스트 1.1M
GPT Luna Latest

This model always redirects to the latest model in the GPT Luna family.

텍스트컨텍스트 1.1M
GPT Image 2.5 Sunburst

GPT Image 2.5 Sunburst is an image generation and editing model from OpenAI, positioned as the precision-oriented tier of the GPT Image 2.5 series. It is suited to detailed creative work where editing accuracy matters more than generation speed, via the dedicated Images API.

이미지US$8 / US$30
GPT Image 2.5 Flare

GPT Image 2.5 Flare is an image generation and editing model from OpenAI, positioned as the speed-oriented tier of the GPT Image 2.5 series. It is suited to high-volume everyday generation, creator content, and rapid prototyping via the dedicated Images API.

이미지US$8 / US$30
GPT-6 Astra

GPT-6 Astra is OpenAI's flagship model for demanding end-to-end work. It is suited for advanced analysis, software engineering, deep research, scientific work, and document creation, with particular strengths in long-horizon agentic tasks that involve computer and browser use.

텍스트컨텍스트 1.1MUS$10 / US$50