MModelspectraIndependent AI Model Intelligence
Model profile · Updated 2026-09-08
Model Profile · Alibaba · CN

Qwen3.8-Max

Best forChinese tasks, Multimodal, Coding
LicensingClosed API
Updated2026-09-08
01

Snapshot

Overall
82
rank #4 of 22
Coding
77
rank #6 of 22
Multimodal
92
rank #2 of 22
Input / Output
$2.5 / $7.5
USD per 1M tokens
Context
1M
usable recall 95
Speed
~55 tok/s
TTFT 0.6s

Aggregated from public sources and independently weighted; methodology on the Terms page. Scores within 3 points are treated as statistically tied.

02

Specs & pricing

Every tracked field. List API prices in USD per 1M tokens; verify current vendor pricing before purchase.

VendorAlibaba (CN)
Released / current2026.07
LicensingClosed / managed API
Context window1M
Max output128K
Effective-context score95
Input / output price$2.5 / $7.5 per 1M tokens
Cache discountPrompt cache 80% off
Free tierFree credits for new Alibaba Cloud Bailiang users
Speed / TTFT~55 tok/s / 0.6s
Function calling88
Refusal rate~10%
English / Chinese78 / 98
Modalitiestext, image
Fine-tuningYes
Private deploymentYes (enterprise)
SOC2 / no-trainno / yes
03

Strengths & watch-outs

Strengths

  • Top-tier Chinese
  • Balanced multimodal
  • Alibaba Cloud ecosystem
  • Good value

Watch-outs

  • Average English reasoning
  • Restricted international access
04

Where it ranks among 22 models

Overall 82/100 (#4), coding 77 (#6), multimodal 92 (#2). See the full boards and side-by-side compare on the leaderboard.
05

Pricing in practice

Illustrative monthly bill for 100M input + 30M output tokens: $475 at list; about $295 with 90% of inputs cache-hit. Your mix and cache-hit ratio change this.

Illustrative model using public list rates before any enterprise agreement; cache discount: Prompt cache 80% off.

06

Frequently asked questions

How much does Qwen3.8-Max cost?
List API pricing is $2.5 input and $7.5 output per 1M tokens. Prompt cache 80% off. Free credits for new Alibaba Cloud Bailiang users
What is Qwen3.8-Max's context window, and is it usable end to end?
It advertises a 1M window with 128K max output; its effective-context score is 95/100, which is the better predictor of whether long-document details are actually retained.
Is Qwen3.8-Max good for coding?
Its coding aggregate is 77/100, rank #6 of 22. It is a strong choice for most product engineering.
Can I self-host Qwen3.8-Max?
No. It is a closed managed API (private deployment may be negotiated for enterprise; there is no weights download.
How strong is it in Chinese and on multimodal inputs?
Chinese score is 98 versus English 78; multimodal aggregate is 92/100 (rank #2), accepting text, image.