News for hackers. Stuff that matters.
Re:Mercury 2: The fastest reasoning LLM, powered by diffusion by dvt on Tuesday February 24, 2026 @07:09PM (#47145428)
What excites me most about these new 4figure/second token models is that you can essentially do multi-shot prompting (+ nudging) and the user doesn't even feel it, potentially fixing some of the weird hallucinatory/non-deterministic behavior we sometimes end up with.
Re:Mercury 2: The fastest reasoning LLM, powered by diffusion by volodia on Tuesday February 24, 2026 @09:17PM (#47146468)
That is also our view! We see Mercury 2 as enabling very fast iteration for agentic tasks. A single shot at a problem might be less accurate, but because the model has a shorter execution time, it enables users to iterate much more quickly.
Re:Mercury 2: The fastest reasoning LLM, powered by diffusion by lostmsu on Tuesday February 24, 2026 @10:41PM (#47147033)
Regular models are very fast if you do batch inference. GPT-OSS 20B gets close to 2k tok/s on a single 3090 at bs=64 (might be misremembering details here).
Re:Mercury 2: The fastest reasoning LLM, powered by diffusion by rahimnathwani on Wednesday February 25, 2026 @01:59AM (#47148258)
Right but everyone else is talking about latency, not throughput.