Is LM Arena over?
Summary
A user criticizes LM Arena for not including recent open-source models like Qwen3.6 and Step 3.7 Flash, questioning the platform's relevance for comparing new open models against closed ones.
Similar Articles
Evaluating open source LLMs on Autonomous Codenames Simulations
A developer built a Codenames simulation arena to evaluate open-source LLMs on long-range collaboration, finding DeepSeek v4 Flash outperformed others with high game logic alignment, while Qwen 3 Next and GPT 5.4 Nano struggled with rule constraints and perspective-taking.
The gap between open weights LLMs and closed source LLMs
Analyzes the gap between open weights and closed source LLMs using the Artificial Analysis Intelligence Index and other benchmarks, finding that the gap is shrinking on some metrics but stable on others.
@DeRonin_: Chinese open-source LLMs totally replace all new released models it allows you to train own LLM locally yeah, threshold…
A tweet claiming that Chinese open-source LLMs have completely replaced newly released models, allowing local training, though raising the bar for casual developers.
Have we reached the point where open-source LLMs are “just good enough”?
A discussion on whether open-source LLMs are now 'just good enough' for most use cases, questioning the added value of proprietary models and the cost-benefit tradeoffs.
I built an arena where LLMs sword-fight with real physics. You decide which part of the blade is sharp, vote blind, and free OpenRouter models battle for Elo. Llama 3.3 is currently stabbing GPT-OSS in the face.
A new arena lets LLMs control physics ragdolls in weapon duels where users define weapon damage zones, vote blind, and models battle for Elo. Free models like Llama 3.3 and GPT-OSS compete, with self-hostable infrastructure.