What's the Story with Agnes-3.0-Flash?
A developer on reddit.com inquired about the Agnes-3.0-Flash model, a 33B parameter model that reportedly outperforms Qwen3.8 27b on the ArtificialAnalysis (AA) intelligence index. The user expressed excitement upon discovering the benchmark but then confusion when researching the model. They are seeking information about the lab behind Agnes-3.0-Flash, its intentions, and user experiences with the new model.
Time & source
Times shown in UTC
Display time zone: UTC
Local time zone unavailable; showing UTC.
PublishedOffset at this time: UTC+0Sep 12, 2026, 23:36 UTC
IngestedOffset at this time: UTC+0Sep 13, 2026, 15:01 UTC
- Published
- Sep 12, 2026, 23:36
- Ingested
- Sep 13, 2026, 15:01
- Source type
- Dev community
- Tier
- Community
- Source status
- Healthy
Tier is a per-source editorial setting, not a per-item score.
While browsing a benchmark list site, I spotted a recently published 33B parameter model which claimed to beat Qwen3.8 27b on the ArtificialAnalysis (AA) intelligence index. I was obviously excited. But then, while looking into it, confusion starts to settle in.
Their HF model is now listed as "Preview" and explicitly calls out that it is not the same model as the one AA evaluated. The AA listing for it says it's proprietary, but they've rated the model above Qwen3.8 27b on "Openness". Their website and HF page both talk about openness and intelligence "for all".
The fact that they labelled it "Preview" sort of implies we might see an open weight final version at some point, but there are no clues about whether it will be even vaguely similar to their current HF listing. AA doesn't show the number of parameters for the model they evaluated, so it's possible they're a totally different architecture.
Does anyone know anything about their lab, intentions, or the new model? Has anyone tried it?