I can see this pattern of many people using Qwen 3.8 27B for local inference both on Apple Silicon and x86. This implies the model must be very good, given all these peoples' opinion converges on it.
If you're looking for a coding agent that would fit nicely into resource-constrained environments (such as laptops, or tiny VPS servers, or tiny single-board computers, etc), and would also work great with local models - you might also like hax (https://usehax.dev/). 0.7 MB dynamically linked native C binary, few MBs of RAM usage when running, auto-discovers config from running local llama-server, and uses minimalist system prompt and tools for lean context usage.
HN seems to have had a stream of agent harness benchmarks floating past. And every time I wonder where the people who create these tests are looking when they're deciding which harnesses to test? Because right now nobody seems to bother testing mine! (https://juggler.studio)
I know Juggler's very new, but there's so much churn going on in this area that it's hard to know where I should be pushing it. It's hard to guess whether juggler's strengths would played well with a particular test like this, or made it look bad, all feedback about the kind of parameters people are interested in is useful to know when I'm deciding what to optimise.
The same thing as the last word of "That is the difference between 22 and 226 seconds, measured." Techies I know would mostly omit "measured"; the rest would show, not tell.
A bit off topic because I'm not using local models, but I recently benchmarked codex vs pi vs omp with my workload and found codex to be both faster and more token efficient than pi/omp. There was not a single case for which pi was faster/cheaper
Notion is a "knowledge database" with awful performance and jank, that some people have decided sounded like a perfect place to host their blog for whatever reason. But these shared pages been as buggy as the first time I saw them years ago, not sure what they're doing.
Nice writeup! I imagine these results change as harnesses are updated, so you'd need to frequently rereview.
I'd love to see a tiny, reproducible benchmark repo that anyone can drop on their own hardware and then run against all harnesses at once to compare the per turn prefix token count, time to the first token, experienced tokens/sec (and prefill), cache reuse % and a pass rate on a deterministic set of small tasks. I think it could also be useful to have some way to share results and hardware for others to compare.
Fun reference I tested on 32 GB ram laptop with no extra GPU: llama.cpp: “what is ls”, almost immediate starts answering at one ~word/sec. Ask opencode with same model (some gwen e4b or something) to check what’s in its working directory: 20 min to response.
Opencode system prompt contains a lot of stuff but even worse is oh-my-pi where their long prompt looks like random garbage hallucinated by a 2023 LLM:
“Chad” initially looked interesting but the minute I saw the ai-written markdown and giant commit I just left. I just can’t bring myself to read someone elses’ slop, regardless of performance.
If all a developer hand writes is a truthy and readable markdown document, I really don’t care if the rest of the project is vibe coded, but I struggle to get interested in AI generated summaries and docs.
I can see this pattern of many people using Qwen 3.8 27B for local inference both on Apple Silicon and x86. This implies the model must be very good, given all these peoples' opinion converges on it.
If you're looking for a coding agent that would fit nicely into resource-constrained environments (such as laptops, or tiny VPS servers, or tiny single-board computers, etc), and would also work great with local models - you might also like hax (https://usehax.dev/). 0.7 MB dynamically linked native C binary, few MBs of RAM usage when running, auto-discovers config from running local llama-server, and uses minimalist system prompt and tools for lean context usage.
ive looked at your project before but forgot about it. I think mine is in a similar vein: https://github.com/mischief/clm
it grew out of annoyance of dependencies on js runtimes, probably similar to you. mine additionally works on solaris and esp32.
could be interesting to collaborate!
HN seems to have had a stream of agent harness benchmarks floating past. And every time I wonder where the people who create these tests are looking when they're deciding which harnesses to test? Because right now nobody seems to bother testing mine! (https://juggler.studio)
I know Juggler's very new, but there's so much churn going on in this area that it's hard to know where I should be pushing it. It's hard to guess whether juggler's strengths would played well with a particular test like this, or made it look bad, all feedback about the kind of parameters people are interested in is useful to know when I'm deciding what to optimise.
What is this supposed to mean?
"it spreads up to 50% between nights, so nothing between the lean arms is a finding."
The same thing as the last word of "That is the difference between 22 and 226 seconds, measured." Techies I know would mostly omit "measured"; the rest would show, not tell.
Pangram fires as usual. Human opening, 75% machine.
Whatever it is, it’s just as hilarious as Engrish…
Means Claude can't write for shit.
Or chad
A bit off topic because I'm not using local models, but I recently benchmarked codex vs pi vs omp with my workload and found codex to be both faster and more token efficient than pi/omp. There was not a single case for which pi was faster/cheaper
Pi is a very basic harness by design. On the other hand OMP is a bloated mess of other people’s workflows.
The trick with pi is to extend it yourself as you use it. It’s pretty easy to do.
Every single article and banchmark say pi saves token by default, the more I add extensions the more token hungry it gets.
pi used 2-3x the tokens of codex. pi with subagent pkg used 8x-10x the tokens of codex.
I don't see how adding bloat to pi would make it more token efficient if the baseline is so poor to start with
What is with the website though? Rubbish scrolling. Junky rendering with artifacts if you scroll fast.
Notion is a "knowledge database" with awful performance and jank, that some people have decided sounded like a perfect place to host their blog for whatever reason. But these shared pages been as buggy as the first time I saw them years ago, not sure what they're doing.
Nice writeup! I imagine these results change as harnesses are updated, so you'd need to frequently rereview.
I'd love to see a tiny, reproducible benchmark repo that anyone can drop on their own hardware and then run against all harnesses at once to compare the per turn prefix token count, time to the first token, experienced tokens/sec (and prefill), cache reuse % and a pass rate on a deterministic set of small tasks. I think it could also be useful to have some way to share results and hardware for others to compare.
I've made https://maki.sh for use cases such as this
Fun reference I tested on 32 GB ram laptop with no extra GPU: llama.cpp: “what is ls”, almost immediate starts answering at one ~word/sec. Ask opencode with same model (some gwen e4b or something) to check what’s in its working directory: 20 min to response.
Opencode system prompt contains a lot of stuff but even worse is oh-my-pi where their long prompt looks like random garbage hallucinated by a 2023 LLM:
https://m.youtube.com/watch?v=c_fQoDkULl0 (see around 8:00)
Neat article.
“Chad” initially looked interesting but the minute I saw the ai-written markdown and giant commit I just left. I just can’t bring myself to read someone elses’ slop, regardless of performance.
If all a developer hand writes is a truthy and readable markdown document, I really don’t care if the rest of the project is vibe coded, but I struggle to get interested in AI generated summaries and docs.
I for one am excited to learn more about how it spreads up to 50% between nights, and how nothing between the lean arms is a finding.
Don't forget it's, measured.
wtf is wrong with scrolling on that website?