Lemonade Server by AMD
Hacker ApprovedRanked 8 of 27 local LLM runners by recommenders
Free AMD-backed server running local models on AMD GPUs and NPUs, handling backends and OpenAI-compatible endpoints. Main caveat: proprietary NPU kernels.
Free AMD-backed server running local models on AMD GPUs and NPUs, handling backends and OpenAI-compatible endpoints. Main caveat: proprietary NPU kernels.
36 people gave an opinion
- recommend it
- 30
- criticize it
- 6
- approval share
- 83%
Recommended in 11 of 13 months
- November 2025: recommended
- December 2025: not recommended
- January 2026: recommended
- February 2026: recommended
- March 2026: recommended
- April 2026: recommended
- May 2026: recommended
- June 2026: recommended
- July 2026: recommended
- August 2026: recommended
- September 2026: recommended
- October 2026: recommended
- November 2026: not recommended (month in progress)
Why it carries the badge
- 30 people recommend itNeeds 10 or more, met.
- 5 recommenders for every criticNeeds 2 to 1 or better, met.
- Recommended in 11 of 13 monthsNeeds 4 or more, 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
Easy setup and installation
“I was surprised at how easy it was to install on my Ubuntu host!”
Comment from 11 Sep 2026 in “Easy Local Copilot with VS Code and Lemonade” “LM Studio or Lemonade and you are set up in minutes”
Comment from 3 Aug 2026 in “Qwen3.8-Max: A New Bar for Coding and Cowork” “Just go download and install lemonade server.”
Comment from 30 Oct 2026 in “Fiber lasts decades, GPUs last three years: the AI boom isn't the dot-com boom [video]” - Comment from 29 Jun 2026 in “Qwen 3.6 27B is the sweet spot for local development”
- Comment from 2 Apr 2026 in “Lemonade by AMD: a fast and open source local LLM server using GPU and NPU”
Optimized for AMD hardware
“It’s an easy way to get started and maintain a local AI stack that concentrates on AMD optimization.”
Comment from 2 Apr 2026 in “Lemonade by AMD: a fast and open source local LLM server using GPU and NPU” “an excellent runtime to serve local AI models through AMD NPUs and GPUs”
Comment from 30 May 2026 in “It's hard to justify buying a Framework 12” “in particular if you are running AMD hardware due to extra optimization”
Comment from 1 Jun 2026 in “Odysseus – self-hosted AI workspace” - Comment from 2 Apr 2026 in “Lemonade by AMD: a fast and open source local LLM server using GPU and NPU”
- Comment from 2 Apr 2026 in “Lemonade by AMD: a fast and open source local LLM server using GPU and NPU”
One server handles many backends
“Having one server handle all of that behind OpenAI-compatible endpoints is a real quality of life improvement”
Comment from 2 Apr 2026 in “Lemonade by AMD: a fast and open source local LLM server using GPU and NPU” “The multi-models on Lemonade make it straight forward to run a stack for LLM”
Comment from 15 Jun 2026 in “Ask HN: Has anyone replaced Claude/GPT with a local model for daily coding?” “because it does a great job of handling backends and providing a server.”
Comment from 27 Oct 2026 in “Friendly UI and server and agent for local LLMs” AMD-backed official project
“A team from AMD maintains Lemonade, another all-in-one setp with convenient installers for setting everything up”
Comment from 12 Aug 2026 in “llama.cpp” “The fact that AMD is backing an official inference server that handles the driver/dependency maze is huge.”
Comment from 2 Apr 2026 in “Lemonade by AMD: a fast and open source local LLM server using GPU and NPU” “officially supported project Lemonade”
Comment from 19 Apr 2026 in “My first impressions on the ROCm and Strix Halo”
What people warn about
Each mark is one comment
NPU code is proprietary
“Lemonade Server uses NPU low-level code that's proprietary, not available as open source.”
Comment from 19 Mar 2026 in “Nvidia greenboost: transparently extend GPU VRAM using system RAM/NVMe” “the NPU models/kernels this uses are proprietary and not available as open source”
Comment from 2 Apr 2026 in “Lemonade by AMD: a fast and open source local LLM server using GPU and NPU”
Other local LLM runners people recommend
The four most recommended other local LLM runners, 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 |
|---|---|---|---|---|---|---|
| llama.cpp | 471 | 109 | 81% | 13 | 1 yr4 owners | Approved |
| Ollama | 350 | 247 | 59% | 12 | 1.5 yrs8 owners | — |
| LM Studio | 217 | 74 | 75% | 12 | 3 yrs4 owners | Approved |
| DS4 by antirez | 109 | 20 | 84% | 7 | 1 mo3 owners | Approved |
| Lemonade Server by AMDThis page | 30 | 6 | 83% | 11 | — | Approved |
What people said
40 of the 49 comments that mention it, newest first. Each links to the original comment.
“Just go download and install lemonade server.”
“because it does a great job of handling backends and providing a server.”
“you'd just download lemonade server and use models designed explicitly for the hardware”
“Also supports Lemonade and Ollama, which are.”
“I was surprised at how easy it was to install on my Ubuntu host!”
“works great with their lemonade-server”
“Lemonade has industry-leading loss ratios now”
“A team from AMD maintains Lemonade, another all-in-one setp with convenient installers for setting everything up”
“Lemonade-server works pretty well (most of the time). It wraps llama.cpp and other runtimes”
“LM Studio or Lemonade and you are set up in minutes”
“it will tell you exactly what will fit”
“The UI has improved a lot, but be careful as the most of the models default to very small 4K context windows by default.”
“Highly recommend lemonade server if you have a strix halo desktop”
“Lemonade + qwen3.6 35B MTP @ Q8_0 + anythingLLM (in docker) replaced 90%+ of my AI usage.”
“lemonade is just an easier to manage system”
“Lemonade really made it plug-and-play.”
“I started to experiment with locale LLMs, through ollama and Lemonade.”
“The multi-models on Lemonade make it straight forward to run a stack for LLM”
“in particular if you are running AMD hardware due to extra optimization”
“an excellent runtime to serve local AI models through AMD NPUs and GPUs”
“I hear good things about the Lemonade team”
“Lemonade builds of llama.cpp with Unsloth or Bartowski quants have proven to be excellent.”
“officially supported project Lemonade”
“Having one server handle all of that behind OpenAI-compatible endpoints is a real quality of life improvement”
“the unified api is interesting, but i've found that 'openai compatible' can be leaky”
“Lemonade is the real deal and very good, been using about a year now.”
“So it seems faster in this very limited test.”
“The fact that AMD is backing an official inference server that handles the driver/dependency maze is huge.”
“if there are tools that do all the hard work of making AI tools work well on my hardware, I'm all for it”
“It’s an easy way to get started and maintain a local AI stack that concentrates on AMD optimization.”
“it’s really the easiest method”
“I actually don't like their combined server”
“Really recommend this for AMD hardware!”
“Cool but is there a reason they can't just push PRs to vLLM and llama.cpp?”
“the NPU models/kernels this uses are proprietary and not available as open source”
“Lemonade or even llama.cpp are much better optimised and arguably just as easy to use.”
“Lemonade Server uses NPU low-level code that's proprietary, not available as open source.”
“If you're on an AMD platform Lemonade added image generation in version 9.2”
“use the prebuilt lemonade nightlies”
“AMD’s open-source Lemonade stack + Ryzen AI Max series”
Counts come from 1,795,640 Hacker News comments, November 2025 to November 2026. How we count. Report a mistake.