# Qwen3.8-Max

Alibaba, released 2026-07-19

**Verdict: pending.** Qwen3.8-Max is Alibaba's 2.4-trillion-parameter flagship, previewed July 2026 as a multimodal MoE model with a 984K context window. It is not yet GA: no per-token API pricing, no benchmark table, no open weights, and no model card have been published, and the "second only to Fable 5" claim is Alibaba's own unverified self-assessment. The preview is worth testing for coding and long-context agent workloads at $6/month entry cost, but wait for published benchmarks, per-token pricing, and the promised open-weight release before committing. Last verified 2026-08-03.

## Pricing

- Input: $2.00 / 1M tokens
- Output: $6.00 / 1M tokens
- Context: Model Studio (Singapore); one tier up to 1M tokens

## Strengths

- Published per-token rate: $2 input / $6 output per 1M on Alibaba Cloud Model Studio (Singapore), one tier up to 1M tokens, with context-caching discount
- Native multimodal input — Alibaba Cloud's model list places it under both text generation and image and video understanding
- OpenAI and Anthropic API-compatible — works with Cursor, Claude Code, OpenCode, and Codex
- Choosing on benchmarks — we found no published benchmark table or model card, and the 'second only to Fable 5' claim is Alibaba's own self-assessment
- Self-hosting — 2.4T parameters would need datacenter-scale infrastructure, and we could not confirm an open-weight release
- Regulated environments requiring auditable model cards and verified benchmark data

## Sources

- [Alibaba Cloud Model Studio model pricing](https://www.alibabacloud.com/help/en/model-studio/model-pricing)
- [Kie.ai — What Is Qwen3.8-Max? Alibaba's 2.4T Flagship](https://kie.ai/blog/what-is-qwen3-8-max)
- [MarkTechPost — Alibaba Previews Qwen3.8-Max](https://www.marktechpost.com/2026/07/19/alibaba-previews-qwen3-8-max-a-2-4-trillion-parameter-multimodal-model-days-after-moonshots-kimi-k3-open-weight-launch/)
- [Alibaba Cloud Model Studio: Models](https://www.alibabacloud.com/help/en/model-studio/models)
- [Coursiv — Qwen 3.8: Specs, Pricing, Access & Benchmarks](https://coursiv.io/blog/qwen-3-8)
- [Lumien — Alibaba's Qwen3.8-Max: 2.4 Trillion Parameters, No Benchmarks Yet](https://lumienai.com/news/alibaba-qwen3-8-max-preview-2-4-trillion-parameter-multimodal-model)

Canonical: https://neomanex.com/models/qwen3-8-max