AlibabaProprietary

Qwen3.7-Plus-Preview

Compare this model

Qwen3.7-Plus-Preview is an inference model developed by Alibaba. Equipped with an extensive 256K context window, it provides advanced reasoning capabilities.

Parameters

Undisclosed

Context Window

256K

License

Proprietary

Release Date

2026-05-20

API Pricing

API pricing for this model is not yet available

Strengths

  • Equipped with advanced reasoning
  • Long 256K context processing
  • Latest model architecture

Weaknesses

  • Non-open-source license
  • Restricted, closed environment
  • Undisclosed detailed technical specifications

Use Cases

  • Complex logical reasoning tasks
  • Analysis of long documents
  • Automation of advanced problem-solving

Deep Analysis

Arena Elo

1475

#3 overall in Text Arena (Qwen3.7-Max-Preview ranked #13 overall at 1475)

SWE-Bench Verified

80.4%

vs Claude Opus 4.7: ~85% (estimated); GPT-5.5: ~82% (estimated)

Input Price

$15/1M

Premium tier (similar to Qwen3.6 Max at $1.30/$7.80 per million input/output tokens)

Context Window

256K tokens

Multimodal (vision + text), vs Max-Preview at 1M tokens (text-only)

Vision Ranking

#16 overall

Vision Arena global rank, making Alibaba the #5 lab in vision

Strengths

  • Strong multimodal capabilities with vision input support, excelling in image understanding and OCR tasks
  • Balanced performance across reasoning, coding, and math with competitive benchmark scores
  • Chinese language and cultural context optimization, with high fluency in Chinese text generation

Weaknesses

  • Preview model with no open weights yet, limiting local deployment and fine-tuning options
  • Lower factual recall accuracy with higher abstention rate (48% attempt rate on AA-Omniscience)
  • Currently locked into thinking mode with web search and code interpreter disabled in preview

Competitor Comparison

ModelArenaSWEGPQAPrice
Claude Opus 4.71480~85%92.4%$15/$75
GPT-5.51485~82%93.6%$5/$30
Gemini 3.5 Flash1470~78%91.3%$1.50/$9

Qwen3.7-Plus-Preview is Alibaba's high-performance multimodal reasoning model, positioned as the balanced counterpart to the text-only Qwen3.7-Max-Preview. Released in May 2026 as part of the Qwen 3.7 series, it focuses on reasoning and logical expression while supporting both text and image inputs. The model achieved #16 overall in Vision Arena and #3 in Text Arena rankings, demonstrating strong capabilities across multiple domains including math (#7), software/IT (#9), and expert prompts (#9). As a preview model, it represents Alibaba's continued advancement in the competitive AI landscape, though it remains proprietary with no open weights available. The model is particularly strong in Chinese language contexts and cultural understanding, offering a viable alternative to Western frontier models for developers seeking balanced multimodal capabilities at competitive pricing.

Analysis generated: 2026-05-23