Qwen3.8-Max Review
Truly multimodal: text, images, and video in a single model, at a very friendly price.
Scores
Score by area
Summary
Strengths and weaknesses
Strengths
- AA Index 58 at only $2/$6
- No. 2 on Arena vision, the best image and video handling in its price range
- 1M context window
Weaknesses
- Weights not released, even though earlier Qwen versions were all open
- Behind the top group at agentic coding
- English documentation arrives later than the Chinese version
Review
Detailed review
Ranking source
Updated August 9, 2026 from the Artificial Analysis Intelligence Index (v4.1 scale), cross-checked against each provider's release notes and official pages. Scores on the v4.1 scale can't be compared directly with earlier rankings.
When to use it
Use it for pipelines that involve images and video: writing product descriptions from pictures, pulling content out of clips, and moderating images.
Notes
AA Intelligence Index: 58; context 1M; GA August 3, 2026; price $2/$6; a 2.4T MoE with about 95B active. The best price-to-score ratio among multimodal models.
Ông Chú MMO
AI Reviewer
Reviews are written by the Ông Chú MMO team based on hands-on experience.
Comments
No comments yet.
Similar tools
Compare toolsMultimodal LLM
Muse Spark 1.2
Meta's model that launched alongside the Muse Code terminal agent, with a low price and a score that still sits in the top group.
9.1Multimodal LLM
Gemini 3.1 Pro
Google's highest Pro version, very strong at vision but now lagging in overall score.
8.3Balanced LLM
GPT-5.6 Terra
The mid-tier version of GPT-5.6, scoring close to Sol for just over a third of the price.
9.1
