Qwen3.8 Flash Next, Swift 1.5, and MiMo V2.6: What I’m Testing Now
The Kaitchup – AI on a Budget is a reader-supported publication.
Proof1 independent outlet
Key points
- In this Weekly Kaitchup, I want to discuss:
- Some updates on what I’m running
- I’ve been running Qwen3.8 Flash Next for the past week on a B300, testing all three thinking modes, low, medium, and xhigh, as well as the model with thinking disabled.
- I can already say that Flash Next behaves very similarly to Qwen3.8 27B, but with stronger overall accuracy:
Sources (1)
- [1]Qwen3.8 Flash Next, Swift 1.5, and MiMo V2.6: What I’m Testing NowThe Kaitchup (Benjamin Marie) · Oct 3, 02:00 AM
The Kaitchup – AI on a Budget is a reader-supported publication.
In this Weekly Kaitchup, I want to discuss:
Extractive summary: sentences quoted from the sources.
Before this
- Oct 2, 2026alesha-pro/Qwen3.8-Flash-Next-abliterated-GSQ-RCO-Strata-GGUF
- Oct 1, 2026unslothai/unsloth v0.1.902-beta: Command Palette + Desktop UI/UX
- Sep 30, 2026Qwen3.8 Flash Next GGUF Benchmark: Q4 to Q1 Accuracy and Token Efficiency
- Sep 28, 2026unslothai/unsloth v0.1.900-beta: Laya Decision Models + Library
- Sep 28, 2026Holo4: powering generalist computer-use agents
- Jun 9, 2026Introducing Gemma 4 12B: a unified, encoder-free multimodal model
