Qwen 3.8 27B scores 52 on the Artificial Analysis Intelligence Index
- ID
- 15022
- Status
- summarized
- Published
- 18 Aug 2026, 7:58 AM
- Fetched
- 19 Aug 2026, 6:08 AM
- Provider
- Simon Willison
- Category
- developer-ai
- Original URL
- https://simonwillison.net/2026/Aug/17/qwen-38-27b-scores-52/
- Source URL
- https://simonwillison.net/atom/everything/
Summary
- Score
- 7.5
- Created
- 19 Aug 2026, 6:08 AM
- Tags
- Audience
- developersvibe_codersai_ml_learnersai_agent_users
What happened
Qwen 3.8 27B scores 52 on the Artificial Analysis Intelligence Index, matching GPT-5.6 Luna (max) and trailing GLM-5.2 (max) and DeepSeek V4 Pro 0813 (max) by just one point—despite being 27B parameters versus 753B for GLM and unknown-but-larger for Luna. Simon Willison calls it 'truly astonishing,' though a prior post notes it 'defaults to wildly overthinking things.'
Why it matters
A 27B model matching frontier proprietary models on a standard index means you can potentially self-host or run locally a model competitive with GPT-5.6-class APIs, cutting inference costs dramatically. Before deploying, test whether the 'overthinking' default inflates latency or token costs for your use case.
Discussion angle
Is a 27B self-hosted model now the default choice for Malaysian builders facing API costs and data-residency constraints, or do the overthinking latency and operational overhead still make API calls the pragmatic pick?