Multimodal Benchmark Coding Agents Qwen3.8-Max

Qwen3.8-Max-Preview

First Available on Token Plan

Alibaba Cloud's Qwen3.8-Max-Preview is a preview model release and the exclusive debut model of the Model Studio Token Plan for Individual. It combines frontier-level text reasoning with vision understanding for coding, data analysis, office workflows, and multimodal inputs.

Pricing

Model profile · Current snapshot

Input / 1M

$2.00

Output / 1M

$6.00

Blended / 1M

$3.00

Output TPS

38.42

TTFT (s)

1.73

Model Intelligence

Arena ranking

Benchmarkable

Yes

Model level

release

Intelligence Index

46.9

Coding Index

71.8

GPQA

0.93

HLE

0.43

SciCode

0.53

LCR

0.78

Recent stories

1 linked story

releasePRIMARY2026-08-09

Alibaba releases Qwen3.8-Max with 2.4T MoE and 1M context

Alibaba released Qwen3.8-Max, described in a launch thread as a 2.4T sparse MoE with 95B active parameters, 1M context, native vision/text, and agent benchmarks. API pricing is listed at $2/$6 per 1M tokens, with open weights planned for Hugging Face.