HNHacker News
TopNewBestAskShowJobs

Bnjoroge

321 karma · joined April 13, 2020

submissionscomments
Bnjoroge··on Launch HN: Prized (YC S26) – Let non-engineer staff build secure internal tools
Demo looks cool, congrats on the ship! Probably missed but is it possible to externally share/view the built artifact outside of prized?
Bnjoroge··on Qwen 3.8
i'll take that 100% over llm slop.
Bnjoroge··on Qwen 3.8
what exactly do they do to "tune" it? It's not Deepseek's official harness, and going direct with deepseek using ur own harness is stupid cheap
Bnjoroge··on Qwen 3.8
I use these guys. https://crof.ai/tos. Their prices match deepseek, pretty dang fast, and support ZDR. Maybe they serve quantized models but I havent seen a drop in quality from my evals using them vs Deepseek or even for other models.
Bnjoroge··on G# – A modern .NET language with Go, Kotlin, and Swift ergonomics
It’s a pretty safe assumption that most recent projects use AI in some capacity. Models have gotten pretty good
Bnjoroge··on TS-2026-009: Insecure argument handling in Tailscale SSH permitted root access
Did you try Headscale? https://github.com/juanfont/headscale or netbird? The latter has been great for me.
Bnjoroge··on Show HN: Learn by rebuilding Redis, Git, a database from scratch
I’ll take that over something AI generated that’s not free
Bnjoroge··on SWE-1.7 Reach Near GPT 5.5 and Opus Intelligence
Yea I cant convince myself to pay api pricing unless my employer is doing it
Bnjoroge··on SWE-1.7 Reach Near GPT 5.5 and Opus Intelligence
I’ve unfortunately had to temper my excitement with Cognition’s models/products given the amount of unwarranted hype they created with Devin on first release, but hopefully this is good.
Bnjoroge··on Show HN: role-model, a router for hybrid local/cloud AI
Interesting, any benchmarks for how well it works?
Bnjoroge··on Show HN: Smart model routing directly in Claude, Codex and Cursor
Looks cool, but I couldnt find any benchmarks for how well this does. Any links?
Bnjoroge··on The unbearable cheapness of open weight models
It was a rough heuristic for how much more opus/5.5 would presumably cost if you extrapolate from glm5.2 prices. In any case input tokens for 5.4/4.6 are 70-100% more expensive, cached about 3-100%, and output tokens anywhere from 60%-240% more as per all their current api pricing. I highly doubt 5.4/4.6 are that much more expensive to serve given how cheap and commoditized inference has become, and how comparable they are perfomance wise.
Bnjoroge··on The unbearable cheapness of open weight models
You can make a fairly decent assumption by calculating the margin on serving glm 5.2, and adding say 30% extra costs and it still leaves a healthy margin
Bnjoroge··on The unbearable cheapness of open weight models
I dont think that’s accurate. I mean look at how much more expensive frontier closed source models are vs something like glm 5.2 which is just about as good. Serving glm is really cheap, and high margin. Obviously no one knows, just how much their inference costs, but if we assume that opus/gpt are maybe 15-20% more parameters than glm 5.2, then it makes no sense for them to charge almost much much more than glm
Bnjoroge··on Show HN: Bible as RAG Database
Cool. What rag techniques did you use?
Bnjoroge··on Show HN: Pyor – A desktop PR reviewer for the diffs GitHub chokes on
Nicely done! I've been enjoying using it. Thanks for sharing.
Bnjoroge··on Show HN: CommitGate – Automatically scan your commit for vulnerabilities
Personally, I disable pre-commit hooks because they’re annoying and slow me down. Pre-push hook would make more sense. Secondly, i’d ideally want to use my codex/ claude subs for this, not an api key
Bnjoroge··on Ask HN: Do you use Claude Code, Codex, or something else?
Opencode and other harnesses like Pi do support codex and claude(for now) subscriptions, but the point about security makes sense.
Bnjoroge··on Ask HN: Do you use Claude Code, Codex, or something else?
I use OMP. Was on Pi but switched over because I like OMP's subagents implementation. Use 5.5/4.8 for coming up with a very detailed plan, ds4pro/mimo for implementation of easy features, 5.5/4.6 for larger/harder features, and 5.5/4.6 for a review. I use them in onorca.dev(settled here after trying just about any "agent ade". i also sometimes use Zed as well. I'm curious to know if folks use any meta-harnesses?
Bnjoroge··on Ask HN: Do you use Claude Code, Codex, or something else?
How do you a) share context/chat etc with each other? plugin? have them make a file? Why did you opt for either of them instead of something like opencode that supports those models?
Bnjoroge··on Data Compression Explained (2012)
I wonder what if anything has changed ever since this article. Is llm-based compression more mainstream?
Bnjoroge··on The Token Compression Illusion: Why I'm Skeptical of RTK
This post offers virtually no data to back up their objections and reads as LLM-generated for the most part. I
Bnjoroge··on Sogen – High-performance Windows and Linux userspace emulator
Incredibly cool
Bnjoroge··on DeepSeek Introduces Vision
I hope they bring it to their apis, especially v4flash. I find myself using mimo 2.5 more since it supports vision and makes it cheap for doing e2e tests with playwright or similar
Bnjoroge··on US holds off blacklisting DeepSeek, more than 100 firms deemed security risks
Have you tried glm 5.2? It’s better than ds4pro and very close to opus4.6. Alot more expensive than ds4pro/mimo but significantly chesper than opus 4.6 for 90% of the quality
Bnjoroge··on US holds off blacklisting DeepSeek, more than 100 firms deemed security risks
You can BYOK with openrouter and pay nothing up until 1mil requests/month iirc which is pretty generous for one person
Bnjoroge··on US holds off blacklisting DeepSeek, more than 100 firms deemed security risks
Deepseek direct is atleast 2x cheaper than other providers serving it. Maybe their caching strategy is just significantly better or they’re subsidizing api pricing. I think it’s the former. $2 is closer to 50 million v4pro tokens in my experience
Bnjoroge··on US holds off blacklisting DeepSeek, more than 100 firms deemed security risks
What are interesting problems? Deepseek and some other open source models do decently well in the gpu kernel benches. https://kernelbench.com/hard Most people think they’re working on “very hard” problems and they really aren’t
Bnjoroge··on Running local models is good now
You can specify which providers you want to serve your model in OpenRouter. Then you can chose US-based ones.
Bnjoroge··on Iroh 1.0
They should basically replace any Why Iroh vs Tailscale with this. Thank you
← PreviousPage 3 of 9Next →