Someone built a benchmark to answer one question

What happens when Jev picks the tool instead of the LLM? 100 tools. 6 tasks. 8 models. 2 modes. every result logged and visualized. the setup: llm-direct mode - the LLM sees all 100 tools on every step and picks one.

PolyBenderOct 51

The AI Speed Tax: 5 out of 6 Fast Tiers Are Overpriced

The formula for value is simple: Tax = Price Ratio / p50 Speed Ratio. A Tax <= 1.0 means you actually get what you paid for. The clocked numbers: - Qwen 3.8 Max Prime: 2x price / 2.03x speed -> 0.99 Tax (n=753, thin data) -

mojeskoOct 515

I benched every gaming gpu tier from 8 to 24gb and wrote one guide for all of you

So you know exactly what your card can run, then i checked it against steam's 15 most owned gpus and 14 of them run a 27b today 8gb, the 5060, 4060, 3050, 3060 ti, 3070 and the 4060 and 5060

Sudo suOct 518

I still can't get over how close the qwen 3.8 125b landed to the glm 5.3 flash 320b building the same floating tree.

So watch! here is every number from both runs in one place, serve settings included, save it if you're picking a local model for agent builds glm 5.3 flash

Sudo suOct 57