Qwen 3.8 27B Matches GPT 5.6 Luna (max) on Artificial Analysis

Melvin Vivas · X post · 2026-08-18 · Open on X

Topics: LLM Fundamentals, Industry Trends & Job Market · Level: beginner

Summary

On Artificial Analysis's intelligence comparison, Qwen 3.8 27B, a fairly small open-weight model, scores about the same as GPT 5.6 Luna at max reasoning. The post shows how to use independent benchmark sites to compare models.

Key points

• Qwen 3.8 27B scores on par with GPT 5.6 Luna (max) on Artificial Analysis's intelligence comparison.
• A 27B open-weight model now competes with a frontier proprietary model on this index.
• Artificial Analysis compares models on intelligence, performance (speed) and price.
• Check independent leaderboards like this one when choosing a model, rather than relying only on vendor claims.

Resources mentioned

• Artificial Analysis – Comparison of AI Models across Intelligence, Performance, and Price · website · artificialanalysis.ai · free
  Benchmark site comparing AI models on an intelligence index, speed and price, here filtered to the GPT models in Codex.
  Also in: Comparing GPT Models in Codex by Intelligence and Cost per Task (Melvin Vivas on X · notes)
• Qwen3.8-27B · tool · huggingface.co · free
  A 27B-parameter open-weight model from Alibaba's Qwen family that you can run locally or call through hosted APIs.
  Also in: Deploying Qwen3.8 27B on Hugging Face Inference Endpoints (Melvin Vivas on X · notes), Using Hugging Face credits: Jobs, Inference Endpoints and Open Models (Melvin Vivas on X · notes), Running Qwen3.8-27B Locally on an M5 Max MacBook with Inco Splash (Melvin Vivas on X · notes), Ternary Bonsai 2 27B: 9x smaller model keeping 98.2% of benchmark scores (Melvin Vivas on X · notes) and 24 more
• GPT 5.6 Luna · tool · openai.com · paid
  The OpenAI model used inside Codex for the demo. The transcript gives the variant name as 'Soul', which is unclear.
  Also in: Set Codex subagent model and reasoning to save usage limits (Melvin Vivas on X · notes), Use GPT-5.6 Luna in Codex for Terminal Tasks (Melvin Vivas on X · notes), Match Reasoning Effort to Task Length in Codex (Astra/Sol) (Melvin Vivas on X · notes), Run Coworker desktop agents cheaply with GPT-5.6 Luna on OpenRouter (Melvin Vivas on X · notes) and 24 more

Try this

• Use Artificial Analysis to compare models on intelligence, speed and price before choosing one.