Re:$500 GPU outperforms Claude Sonnet on coding benchmarks
by riidom on Thursday March 26, 2026 @08:04PM
(#47537479)
Not a word about the tok/sec, unfortunately.
[ Parent
| on HN ]
Re:$500 GPU outperforms Claude Sonnet on coding benchmarks
by arjie on Thursday March 26, 2026 @09:43PM
(#47538158)
It won’t be meaningful considering the architecture: it’s a harness around the model that generated multiple solutions in multiple passes using the test to measure compliance and repair broken solutions. The resulting program won’t be streamed to you because it has existed for minutes as it goes through the cycle. It’s more for an asynchronous use-case.
I, too, was interested because I am always eager to use local models in my claw-like. It looks like this could be useful for an async portion of the harness but it wouldn’t work in interactive contexts.
Very cool ensemble of techniques, particularly because they’re so accessible. I think I will use this form for reusable portions of web browsing functionality in my personal agent.
Re:$500 GPU outperforms Claude Sonnet on coding benchmarks
by Octoth0rpe on Thursday March 26, 2026 @11:07PM
(#47538695)
> A single patched llama-server runs on K3s, providing both generation with speculative decoding (~100 tok/s)
There seems to be at least some detail on that point.