The Download: Attach images in GitHub CLI, Qwen3.8-Max, AI pull request lingo & more

GitHub shares a quick “The Download” roundup covering a new GitHub CLI capability for attaching images and videos to pull requests, plus several AI-related updates including local coding-agent models, large-scale LLM releases, and benchmarks and analyses focused on AI code review and PR-writing language.

Overview

Attach images and videos to pull requests from the terminal

• GitHub CLI now supports attaching images and videos directly to pull requests from the terminal.

Meta Muse Glimmer 30B for local coding agents

• The episode highlights Meta’s Muse Glimmer 30B, positioned for running local coding agents on 24GB GPUs.
• Resources:
  • https://huggingface.co/meta-models/Muse-Glimmer-30B
  • https://research.meta.ai/static/muse-glimmer-methodology

Alibaba Qwen3.8-Max

• The episode covers Alibaba’s Qwen3.8-Max (noted as a 2.4T parameter model).
• Resource:
  • https://qwen.ai/blog?id=qwen3.8

Benchmarks and discussion: AI code review quality

• The episode points to new benchmarks examining AI code review quality, including BullshitBench.
• Resources:
  • https://petergpt.github.io/bullshit-benchmark/viewer/index.v2.html
  • https://x.com/petergostev/status/2080807184313585711

Project pick: PR vocabulary trends (“load-bearing” vocabulary)

• The episode highlights Louis Abraham’s project analyzing AI vocabulary trends in GitHub pull request descriptions (including references to Claude).
• Resources:
  • https://louisabraham.github.io/load-bearing/
  • https://github.com/louisabraham/load-bearing

Chapters

• 00:00 Welcome to The Download
• 00:32 Attach images and videos with the GitHub CLI
• 01:02 Meta releases Muse Glimmer 30B for local GPUs
• 01:41 Alibaba ships Qwen3.8-Max
• 02:34 Are AI code reviews getting worse?
• 03:33 Project pick: the load-bearing vocabulary of Claude
• 04:29 Outro