AI RACE— The AI Race
AIモデル

Qwen3.7 Plus

Alibaba

試してみる ↗
リリース2026/06
生成速度56 トークン/秒
初動レイテンシ (TTFT)1.58s
入力料金$0.4/1M
出力料金$1.6/1M

ランキング

#82
AIM 47.0最高順位: 第25位
#52
AIM 70.5最高順位: 第16位
#39
AIM 91.3最高順位: 第12位
#84
AIM 80.7最高順位: 第28位

レビュー

Qwen3.7 Plus is a specialized reasoning and coding model, well-suited for engineers and scientists who need strong reasoning capabilities. However, it is not a general-purpose model and is less effective in everyday chat contexts or complex workflow automation.

長所

  • Strong logical reasoning, handling complex problems that require in-depth analysis
  • Solid coding proficiency, capable of debugging and solving demanding technical problems
  • Strong performance in specialized tasks requiring analytical thinking

短所

  • Unremarkable general conversational performance, unsuitable for chatbots or general-purpose interactions
  • Weak in AI agent capabilities and workflow automation, making it a poor choice for building autonomous workflows
  • Falls behind top-tier models in programming, occasionally struggling with complex coding problems

ユースケース

Data analysis, step-by-step reasoning problems, and scientific researchCoding assistance: debugging, code review, and algorithm explanationDemanding technical work requiring in-depth analysis (data engineering, statistics, etc.)

ガイド・動画

Qwen3.7-Plus is a multimodal agent model from Alibaba released in June 2026, supporting text, image, and video input. It is available via the Alibaba Cloud Bailian platform (bailian.aliyun.com) or API at ~$0.4/$1.6 per 1M tokens. Best suited for GUI automation (interface recognition and action), video/screenshot analysis, automated coding, and agentic workflows. Tip: pair with tool-use to run autonomous agent loops — the model excels at long multi-step tasks with its 1M context window.

レビュー記事