Who/Where can I find the most honest and non-biased benchmarks for the models?

<!-- SC_OFF --><div class="md"><p>I’m wondering if instead I should be looking backwards and saying “I like model X, let’s see where it is on benchmarks” and then find others who

<!-- SC_OFF --><div class="md"><p>Does anyone here even rely on external measurements or do they just run their own tests? Personally I just use the model and go by feel because

<!-- SC_OFF --><div class="md"><p>How can I keep up to date on the latest models? Is there a website with the latest releases, benchmarks, etc?</p> </div><!-- SC_ON --> &#32;

<!-- SC_OFF --><div class="md"><p>I've always been under the impression that open models were closely trailing behind closed source models on nearly every benchmark from LM Arena,

<!-- SC_OFF --><div class="md"><p>Qwen 3.8 27b </p> <p>Llama.cpp pp 544 tg 40 Exllama pp 633 tg 95</p> <p>Qwen 3.6 35b a3b</p> <p>Llama.cpp pp 563 tg 84 Exllama pp 2274 tg 150</p>

<table> <tr><td> <a href="https://www.reddit.com/r/LocalLLaMA/comments/1vkvdg0/i_made_a_webdesign_benchmark_for_local_models/"> <img alt="I made a web-design benchmark for local

<!-- SC_OFF --><div class="md"><p>I have been following the akitaonrails coding benchmark which tests against a fixed rails + Rubyllm + docker task rather than vendor-reported