Llmspeed
Ranked 3 of 4 model fit calculators by recommenders
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.
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.
17 people gave an opinion
- recommend it
- 8
- criticize it
- 9
- approval share
- 47%
Recommended in 1 of 13 months
- November 2025: not recommended
- December 2025: not recommended
- January 2026: not recommended
- February 2026: not recommended
- March 2026: not recommended
- April 2026: not recommended
- May 2026: not recommended
- June 2026: not recommended
- July 2026: not recommended
- August 2026: not recommended
- September 2026: not recommended
- October 2026: recommended
- November 2026: not recommended (month in progress)
Why it doesn’t carry the badge
- 8 people recommend itNeeds 10 or more, not met.
- 0.8 recommenders for every criticNeeds 2 to 1 or better, not met.
- Recommended in 1 of 13 monthsNeeds 4 or more, not met.
The same three rules apply to every product, and the badge can’t be bought. How the badge works
What people say about it
The reasons people give, with the number of comments that give each one. Open a reason to read what they said.
Why people recommend it
Each mark is one comment
Useful before buying hardware
“I wished there was such a tool before picking my hardware, it's useful. But the estimates needs to be somewhat correct.”
Comment from 17 Oct 2026 in “Show HN: Llmspeed – feel how fast a local LLM runs before you buy the hardware” “This is great. It solves two of my issues”
Comment from 17 Oct 2026 in “Show HN: Llmspeed – feel how fast a local LLM runs before you buy the hardware”
What people warn about
Each mark is one comment
Estimates often inaccurate
“the site claims qwen 27b to use 2x more memory per token than gemma, which is the opposite from what I know”
Comment from 17 Oct 2026 in “Show HN: Llmspeed – feel how fast a local LLM runs before you buy the hardware” “This is not accurate, I get like 80t/s on Q3.8 27b UD Q4_K_XL on my 4090”
Comment from 17 Oct 2026 in “Show HN: Llmspeed – feel how fast a local LLM runs before you buy the hardware” “Bench is laughably out sync with reality too.”
Comment from 18 Oct 2026 in “Show HN: Llmspeed – feel how fast a local LLM runs before you buy the hardware” - Comment from 17 Oct 2026 in “Show HN: Llmspeed – feel how fast a local LLM runs before you buy the hardware”
Ignores tuning knobs and hybrid GPUs
“There are many knobs to tweak to make it faster, which the site ignores.”
Comment from 17 Oct 2026 in “Show HN: Llmspeed – feel how fast a local LLM runs before you buy the hardware” “Doesn't provide for hybrid setups RTX5070 + RTX5080 for example.”
Comment from 17 Oct 2026 in “Show HN: Llmspeed – feel how fast a local LLM runs before you buy the hardware”
Other model fit calculators people recommend
The three most recommended other model fit calculators, with the same counts for each, so you can compare like for like. The tick on each bar marks the 2 to 1 line the badge needs.
| Product | Recommend | Criticize | Approval | Months | Owned | Badge |
|---|---|---|---|---|---|---|
| llmfit | 20 | 2 | 91% | 6 | — | Approved |
| canirun.ai | 12 | 7 | 63% | 7 | — | — |
| LlmspeedThis page | 8 | 9 | 47% | 1 | — | — |
| LLM model right-sizing tool | 5 | 2 | 71% | 1 | — | — |
What people said
14 of the 17 comments that mention it, newest first. Each links to the original comment.
“Bench is laughably out sync with reality too.”
“There are many knobs to tweak to make it faster, which the site ignores.”
“This tool has very tiny quants available for each model, and defaults to the tiny quants”
“This is not accurate, I get like 80t/s on Q3.8 27b UD Q4_K_XL on my 4090”
“the site claims qwen 27b to use 2x more memory per token than gemma, which is the opposite from what I know”
“This is great. It solves two of my issues”
“It is nice to have simulated confirmation of the M4 Max 128gb limitations”
“It just spits out a fixed answer about RAM.”
“Cool idea but I'd like to also be able to set a custom value of tok/s”
“It would be great if numbers weren't off”
“these were tested with very minimal token usage”
“Doesn't provide for hybrid setups RTX5070 + RTX5080 for example.”
“This is a great idea but I'm far less concerned about output token speed”
“I wished there was such a tool before picking my hardware, it's useful. But the estimates needs to be somewhat correct.”
Counts come from 1,795,640 Hacker News comments, November 2025 to November 2026. How we count. Report a mistake.