본문으로 건너뛰기
알파
  • 모델
  • 순위
  • Ori
영업팀 문의

푸터

OpenRouterOpenRouter
© 2026 OpenRouter, Inc

제품

  • 채팅
  • 순위
  • 벤치마크
  • 앱
  • 탐색
  • 모델
  • Ori
  • 컬렉션
  • 제공업체
  • 도구
  • 요금
  • Business
  • Enterprise
  • 실험실

회사

  • 소개
  • 블로그
  • 고객
  • 채용
    채용
  • 개인정보 보호
  • 서비스 이용약관
  • 신뢰 센터
  • 지원
  • OR과 호환
  • 데이터
  • 브랜드

개발자

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

연결

  • Discord
  • GitHub
  • LinkedIn
  • X
  • YouTube
qwen의 파비콘

Qwen: Qwen3.5 397B A17B

qwen/qwen3.5-397b-a17b

모델 가중치
비교

The Qwen3.5 series 397B-A17B native vision-language model is built on a hybrid architecture that integrates a linear attention mechanism with a sparse mixture-of-experts model, achieving higher inference efficiency. It delivers state-of-the-art performance comparable to leading-edge models across a wide range of tasks, including language understanding, logical reasoning, code generation, agent-based tasks, image understanding, video understanding, and graphical user interface (GUI) interactions. With its robust code-generation and agent capabilities, the model exhibits strong generalization across diverse agent.

모달리티
입력 / 출력 가격
US$0.39 / US$2.341M당
컨텍스트
262K
출시일
2026년 2월 16일
비교
제공업체요금성능가동 시간벤치마크앱활동자주 묻는 질문탐색

제공업체

여러 회사가 동일한 모델을 호스팅합니다. OpenRouter는 선택한 라우팅 모드인 Balanced(가격 + 속도), Nitro(가장 빠름), Floor(가장 저렴함)에 따라 해당 회사 중 한 곳으로 요청을 라우팅합니다.

요금

고객이 이 모델에 실제로 지불하는 평균 가격을 제공업체가 게시한 가격과 나란히 표시합니다. 캐싱과 할인으로 인해 실제 지불 가격은 정가보다 훨씬 낮은 경우가 많습니다.

성능

처리량은 모델의 출력 속도입니다(초당 토큰 수 — 높을수록 좋습니다). 지연 시간은 전체 왕복 시간입니다(낮을수록 좋습니다). TTFT는 첫 토큰까지 걸리는 시간, 즉 출력이 나타나기까지 걸리는 시간입니다(낮을수록 좋습니다).

가동 시간

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

벤치마크

표준화된 평가 점수와 공통 벤치마크 프롬프트에 대한 이 모델 자체의 출력입니다. 백분율이 높을수록 좋으며, 각 순위는 해당 출처에서 평가한 모델 중 이 모델의 위치를 보여 줍니다.

벤치마크 점수 요약: Qwen: Qwen3.5 397B A17B (Artificial Analysis, Design Arena 및 Vals AI)
출처벤치마크점수
Artificial AnalysisQwen3.5 397B A17B (Reasoning) Coding Index48.2
Artificial AnalysisQwen3.5 397B A17B (Reasoning) Agentic Index8.3
Artificial AnalysisQwen3.5 397B A17B (Reasoning) GPQA Diamond89.3%
Artificial AnalysisQwen3.5 397B A17B (Reasoning) HLE29.0%
Artificial AnalysisQwen3.5 397B A17B (Reasoning) IFBench78.8%
Artificial AnalysisQwen3.5 397B A17B (Reasoning) τ²-Bench Telecom95.6%
Artificial AnalysisQwen3.5 397B A17B (Reasoning) AA-LCR77.3%
Artificial AnalysisQwen3.5 397B A17B (Reasoning) τ-Bench Banking13.4%
Artificial AnalysisQwen3.5 397B A17B (Reasoning) GDPval-AA15.0%
Artificial AnalysisQwen3.5 397B A17B (Reasoning) CritPt1.7%
Artificial AnalysisQwen3.5 397B A17B (Reasoning) SciCode44.8%
Artificial AnalysisQwen3.5 397B A17B (Reasoning) Terminal-Bench Hard40.9%
Artificial AnalysisQwen3.5 397B A17B (Reasoning) Terminal-Bench 4.00.0%
Artificial AnalysisQwen3.5 397B A17B (Reasoning) Terminal-Bench 2.151.3%
Artificial AnalysisQwen3.5 397B A17B (Non-reasoning) GPQA Diamond86.1%
Artificial AnalysisQwen3.5 397B A17B (Non-reasoning) HLE19.8%
Artificial AnalysisQwen3.5 397B A17B (Non-reasoning) IFBench51.6%
Artificial AnalysisQwen3.5 397B A17B (Non-reasoning) τ²-Bench Telecom83.9%
Artificial AnalysisQwen3.5 397B A17B (Non-reasoning) AA-LCR64.3%
Artificial AnalysisQwen3.5 397B A17B (Non-reasoning) CritPt0.9%
Artificial AnalysisQwen3.5 397B A17B (Non-reasoning) Terminal-Bench Hard35.6%
Design ArenaQwen3.5 397B A17B Models Arena 3D Elo1158
Design ArenaQwen3.5 397B A17B Models Arena Code Categories Elo1178
Design ArenaQwen3.5 397B A17B Models Arena Data Visualization Elo1175
Design ArenaQwen3.5 397B A17B Models Arena Game Development Elo1140
Design ArenaQwen3.5 397B A17B Models Arena SVG Elo1136
Design ArenaQwen3.5 397B A17B Models Arena UI Component Elo1158
Design ArenaQwen3.5 397B A17B Models Arena Website Elo1189
Vals AIQwen 3.5 397B A17B SAGE85.5%

앱

이 모델에 가장 많은 트래픽을 보내는 공개 앱입니다. 실제 프로덕션 워크로드의 모습을 보여주는 유용한 지표이자 이 모델이 어떤 사용 사례에 가장 적합한지 알려주는 단서입니다.

활동

시간에 따른 이 모델의 토큰 사용량 및 요청 트래픽.

API

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

현재 EU 역내 라우팅을 통해 사용할 수 없습니다. 리전 내 라우팅 문서

더 많은 모델 탐색

비전 기능을 갖춘 AI 모델: 이미지 이해를 위한 멀티모달 LLM컬렉션AI 모델 순위순위
지연 시간 / 처리량
US$0.60US$3.60US$0.1201.07s102 tps
99.53%
US$0.39US$2.34--0.78s53 tps
99.88%
US$0.50US$3.60US$0.3000.79s70 tps
99.50%
US$0.60US$3.60--1.50s45 tps
99.77%
US$0.55US$3.50US$0.2252.22s49 tps
99.26%
US$0.55US$3.50US$0.1103.22s11 tps
94.38%
US$0.45US$3.00US$0.2200.57s68 tps
99.65%
US$0.60US$3.60--1.44s55 tps
86.54%
US$0.55US$3.50US$0.5501.04s51 tps
99.73%
US$0.75US$4.50--1.21s45 tps
98.77%
처리량
102토큰/초

P50, 제공업체 중 최고

지연 시간
0.57초

P50, 최상의 제공업체

AutoExacto 벤치마크
GPQA DiamondTAU-BenchStreamLake86.6%--Phala87.5%80.4%DigitalOcean87.9%79.8%Alibaba Cloud Int.86.9%80.6%auto-routing87.1%79.7%
+5 개 제공업체 더 보기
가동 시간 (3일)모델에 연결할 수 있었습니다. 요청이 제공업체로 라우팅되었습니다.
100.00%
가용성(3일)모델이 제공업체 중 한 곳에서 추론 결과를 반환했습니다. 오류와 빈 응답은 실패로 집계됩니다.
99.53%

지난 3일간 가용성

최근 72시간
가용성 99.53%
3일 전2일 전어제현재

지난 24시간 동안의 가용성

OpenRouter Availability
99.77%
Without Routing
88.64%

업스트림 제공업체에서 오류가 발생하면 요청 필터가 허용하는 경우 정상 상태인 다른 제공업체로 라우팅하여 복구할 수 있습니다. 제공업체별 가동 시간 데이터에 프로그래밍 방식으로 액세스할 때 사용하는 API는 엔드포인트 API입니다. 자세히 알아볼 수 있는 관련 옵션은 부하 분산 및 제공업체 라우팅 옵션입니다.

1.
https://openclaw.ai/의 파비콘
OpenClaw
OpenClaw is an open-source AI agent that connects to your messaging apps and takes real actions on your behalf, from running commands and browsing the web to managing files and sending emails.
947M토큰
2.
https://nousresearch.com의 파비콘
Hermes Agent
Hermes Agent is an open-source, self-improving AI agent by Nous Research that runs persistently with memory across sessions, and builds reusable skills from experience. It comes with 40+ built-in tools, including web search, browser automation, and vision, plus scheduled automations and subagents.
505M토큰
3.
https://pi.dev/의 파비콘
pi
There are many coding agents, but this one is yours.
497M토큰
4.
https://therealcat.ai/flamekeeper의 파비콘
Flamekeeper Memory Spine
신규
358M토큰
5.
https://olympuscloud.ai/의 파비콘
Olympus Code
신규
314M토큰

자주 묻는 질문

The Qwen3.5 series 397B-A17B native vision-language model is built on a hybrid architecture that integrates a linear attention mechanism with a sparse mixture-of-experts model, achieving higher inference efficiency.

Qwen3.5 397B A17B의 비용은 US$0.39/M 입력 토큰 및 US$2.34/백만 출력 토큰입니다.

Qwen3.5 397B A17B의 컨텍스트 윈도우는 262,144토큰입니다. 최대 65,536개의 완성 토큰을 지원합니다.

예. Qwen3.5 397B A17B은 이를 제공하는 10개 제공업체 중 9개에서 함수 호출을 위한 tools와 tool_choice를 지원하며, 도구를 전송하는 요청은 해당 제공업체로 라우팅됩니다. 또한 response_format의 JSON 스키마를 통해 구조화된 출력을 지원합니다.

Qwen3.5 397B A17B은(는) 텍스트, 이미지 및 동영상을(를) 입력으로 받아 텍스트을(를) 반환합니다.

OpenRouter에서 10개의 제공업체가 Qwen3.5 397B A17B을 제공하며, 해당 제공업체는 Alibaba Cloud Int., DeepInfra, Parasail, DigitalOcean, Phala, AtlasCloud, StreamLake, GMICloud 외 2개입니다. 요청은 사용 가능한 최적의 제공업체로 라우팅되고 장애 발생 시 다른 제공업체로 자동 전환되며, 제공업체 라우팅을 통해 제공업체를 고정하거나 제외할 수 있습니다.

Qwen3.5 397B A17B은 2026년 2월 16일에 출시되었습니다.

다음 제작사의 다른 모델: Qwen

Qwen Image 2.1 Turbo

Qwen Image 2.1 Turbo is the fast, lower-cost tier of the Qwen Image 2.1 series from Qwen, for image generation and editing where throughput and price matter more than peak fidelity. See Qwen Image 2.1 Pro for the higher-quality tier.

이미지from US$0.0184/image
Qwen Image 2.1 Pro

Qwen Image 2.1 Pro is an image generation and editing model from Qwen, the higher-quality tier of the Qwen Image 2.1 series. Choose it over Qwen Image 2.1 Turbo when output fidelity matters more than cost or speed.

이미지from US$0.046/image
Qwen3.8 Max Prime

Qwen3.8 Max Prime is a higher-throughput variant of Qwen3.8 Max from Alibaba's Qwen team, served as a separate SKU at a higher price point. It accepts text, image, and video input and returns text, with a 1M-token context window and reasoning enabled by default. Tool calling, structured outputs, and configurable reasoning effort are supported, matching Qwen3.8 Max.

텍스트1M 컨텍스트US$4 / US$12
Qwen3.8 Omni Flash

Qwen3.8 Omni Flash is an omni-modal reasoning model from Alibaba, the first Qwen model built around agentic capabilities with native audio-video understanding. It is suited for audio-video analysis and summarization, video editing and production workflows, audio-video dialogue, coding, knowledge work, and GUI interaction, and it is particularly strong at long-form multimedia tasks that combine speech, sound, and visual context. It also supports two-channel and four-channel spatial audio understanding.

텍스트1M 컨텍스트US$0.15 / US$0.47
Qwen3.8 Max

Qwen3.8 Max 0902 is an updated snapshot of Qwen3.8 Max from Alibaba's Qwen team. It is a 2.4-trillion-parameter mixture-of-experts model that accepts text, image, and video input and returns text, with a 1M-token context window and reasoning enabled by default.

This snapshot is post-trained for coding and agentic work, including multi-step software projects, multi-tool orchestration, and long-horizon task execution. It also targets chart reasoning, document parsing, and multimodal understanding over long documents and extended video. Tool calling, structured outputs, and configurable reasoning effort are supported.

텍스트1M 컨텍스트US$2 / US$6
Qwen3.8 Flash

Qwen3.8 Flash is a multimodal reasoning model from Alibaba. It is suited for coding assistance, agentic workflows, visual understanding, document and codebase analysis, desktop interaction, chart analysis, and long-video analysis.

텍스트1M 컨텍스트US$0.15 / US$0.47
Qwen3 Reranker 0.6B

Qwen3 Reranker 0.6B is the smallest text reranking model from Alibaba Cloud in the Qwen3 Reranker series. Built on the Qwen3 architecture, it evaluates query-document pairs to produce relevance scores for retrieval and RAG pipelines. Supports 100+ languages with instruction-aware reranking. Designed for low-latency, resource-constrained deployments.

재순위화
Qwen3 Reranker 4B

Qwen3 Reranker 4B is a text reranking model from Alibaba Cloud built on the Qwen3 architecture. It evaluates query-document pairs to produce relevance scores for use in retrieval and RAG pipelines. Supports 100+ languages and programming languages, with instruction-aware reranking that allows customizing scoring criteria per task. A compact 4B alternative balancing performance and inference cost.

재순위화
Qwen3.8 27B

Qwen3.8 27B is an open-weight dense vision-language model from Qwen. It is suited for coding, professional workflows, research, multimodal interaction, and long-running agent tasks, with flexible thinking that can be enabled or disabled.

텍스트1.0M 컨텍스트US$0.005 / US$0.251
Qwen3 Reranker 8B

Qwen3 Reranker 8B is a text reranking model from Alibaba Cloud built on the Qwen3 architecture. It evaluates query-document pairs to produce relevance scores for use in retrieval and RAG pipelines. Supports 100+ languages and programming languages, with instruction-aware reranking that allows customizing scoring criteria per task. Offers strong performance on multilingual benchmarks including MTEB, CMTEB, and MMTEB.

재순위화US$0.20/M tokens
Qwen3 ASR 1.7B

Qwen3 ASR 1.7B is an automatic speech recognition model from Qwen. It supports multilingual language identification and transcription across 30 languages and 22 Chinese dialects, with streaming and offline inference plus segment-level and word-level timestamps.

음성 전사US$0.000008/second
Qwen3 ASR 0.6B

Qwen3 ASR 0.6B is a compact automatic speech recognition model from Qwen. It supports multilingual language identification and transcription across 30 languages and 22 Chinese dialects, with streaming and offline inference plus segment-level and word-level timestamps.

음성 전사US$0.000003/second
Qwen3.8 2.4T A95B

Qwen3.8 2.4T A95B is an open-weight sparse mixture-of-experts model from Qwen and the open-weight variant of Qwen3.8 Max, with 95 billion active parameters out of 2.4 trillion total. It is suited for coding, research, complex reasoning, and agentic workflows.

텍스트1.0M 컨텍스트US$2 / US$6
Qwen Image 3 Pro

Qwen Image 3 Pro is an image generation and editing model from Qwen. It supports precise rendering of text and details as small as 10px, along with richer world knowledge compared to previous generations.

이미지from US$0.04/image
Qwen Image 3

Qwen Image 3 is a unified image generation and editing model from Qwen. It supports precise rendering of text and details as small as 10px, along with a richer world knowledge base than previous generations.

이미지from US$0.03/image
Qwen3.8 Max

Qwen3.8 Max (0803) is the August 3, 2026 checkpoint of Qwen3.8 Max, the flagship model in Alibaba's Qwen3.8 series and the general-availability successor to the Qwen3.8 Max Preview. It is a multimodal reasoning model intended for complex reasoning, visual understanding, coding, and agentic workflows. This checkpoint was superseded by Qwen3.8 Max (0902) on September 5, 2026.

텍스트1M 컨텍스트
Qwen3.7 Flash

Qwen3.7 Flash is a vision-language reasoning model from Alibaba. It is suited for multimodal agents, visual coding, search, and computer interaction, with strengths in object recognition, spatial understanding, and real-world visual perception.

텍스트1M 컨텍스트US$0.03 / US$0.13
Qwen-Audio-3.0-TTS Flash

Qwen-Audio-3.0-TTS Flash is Alibaba's fast, cost-efficient text-to-speech model, generating spoken audio from text via the DashScope Speech Synthesizer API.

음성US$15/M characters
Qwen-Audio-3.0-TTS Plus

Qwen-Audio-3.0-TTS Plus is Alibaba's higher-quality text-to-speech model, generating spoken audio from text via the DashScope Speech Synthesizer API.

음성US$20/M characters
Qwen3.7 Plus

Qwen3.7-Plus is a cost-effective model in Alibaba's Qwen3.7 series. It supports text and image input with text output, building on the series' text capabilities with a comprehensive upgrade to its vision-language abilities while retaining full-stack, agent-level intelligence for coding, tool use, and productivity workflows. Its distinguishing trait is multi-modal interactive hybrid agent capability: it can perceive real-world scenes, read screens and interact with GUIs, generate code from visual references, and perform end-to-end navigation within mobile apps.

텍스트1M 컨텍스트US$0.32 / US$1.28