# Llmspeed

8 Hacker News commenters recommended Llmspeed and 9 criticized it, November 2025 to November 2026. It was recommended in 1 of 13 months.

**Verdict:** A web tool that estimates how fast a local LLM would run on given hardware, so you can check before buying. Critics say its speed and memory estimates often don't match real benchmarks.

## Why people recommend it

- Useful before buying hardware (2 comments): "I wished there was such a tool before picking my hardware, it's useful. But the estimates needs to be somewhat correct." ([source](https://news.ycombinator.com/item?id=50128532)) "This is great. It solves two of my issues" ([source](https://news.ycombinator.com/item?id=50130918))

## What people warn about

- Estimates often inaccurate (4 comments): "the site claims qwen 27b to use 2x more memory per token than gemma, which is the opposite from what I know" ([source](https://news.ycombinator.com/item?id=50130953)) "This is not accurate, I get like 80t/s on Q3.8 27b UD Q4_K_XL on my 4090" ([source](https://news.ycombinator.com/item?id=50131514)) "Bench is laughably out sync with reality too." ([source](https://news.ycombinator.com/item?id=50138394))
- Ignores tuning knobs and hybrid GPUs (2 comments): "There are many knobs to tweak to make it faster, which the site ignores." ([source](https://news.ycombinator.com/item?id=50134730)) "Doesn't provide for hybrid setups RTX5070 + RTX5080 for example." ([source](https://news.ycombinator.com/item?id=50128804))

## Alternatives hackers recommend

- [llmfit](https://hackerapproved.com/p/llmfit.md): 20 recommend
- [canirun.ai](https://hackerapproved.com/p/canirun-ai.md): 12 recommend
- [LLM model right-sizing tool](https://hackerapproved.com/p/right-sizes-llm-models-to-your-system-s-ram-cpu-and-gpu.md): 5 recommend

## Cite as

Hacker Approved, "Llmspeed", https://hackerapproved.com/p/llmspeed/ (counts from 1,795,640 Hacker News comments, November 2025 to November 2026).