Development

A 2.4-trillion-parameter model goes open-weight as a chip unit attacks CUDA

development2026-07-19

Claim

Alibaba is releasing Qwen3.8, a 2.4-trillion-parameter model billed as second only to Fable 5, as open weights, bundling the family at 40% off and plugging into rival coding tools, while its chip unit open-sourced its entire SAIL stack to storm CUDA's moat from below.

Reported metric: 2.4T params

Reporting date: 2026-07-19 (not necessarily the event date)

Sources

Attribution and reporting linked by this record; links do not imply independent verification.

Welcome to July 19, 2026 (theinnermostloop.substack.com)

Knowledge graphopen in browser →

Reported in

Welcome to July 19, 2026

Actor

Evidence for

Supersedes

An open-weight model tops a benchmark ahead of every closed model

Referenced by

Welcome to July 19, 2026

Cite or correct this record

Citation formatCitation text

A 2.4-trillion-parameter model goes open-weight as a chip unit attacks CUDA
Innermost Loop KB — secondary record
Reporting date: 2026-07-19
KB permalink: https://nicholsn.github.io/innermost-loop-kb/developments/2026-07-19-a-two-point-four-trillion-model-goes-open
Claim: Alibaba is releasing Qwen3.8, a 2.4-trillion-parameter model billed as second only to Fable 5, as open weights, bundling the family at 40% off and plugging into rival coding tools, while its chip unit open-sourced its entire SAIL stack to storm CUDA's moat from below.
Attributed sources:
- Welcome to July 19, 2026 — https://theinnermostloop.substack.com/p/welcome-to-july-19-2026
This citation does not imply independent verification of linked-source claims.

Suggest a correction — opens a GitHub draft for you to review and submit.