minimaxir

7 ms·

121.
by minimaxir 2mo ago
That is only when using the DeepSeek API directly. OpenRouter has 24 different providers serving it at existing prices.
122.
by minimaxir 2mo ago
Per the announcement tweet, BaseTen was the inference provider which has 20% cache cost that is typical: https://www.baseten.co/library/deepseek-v4-flash-0731/
123.
by minimaxir 2mo ago
DeepSeek V4 Flash 0731 is an open-weights model which means price is determined by competition/invisible hand of the marketplace: https://openrouter.ai/deepseek/deepseek-v4-flash-0731 With the exception of cache c
124.
by minimaxir 2mo ago
Since the x-axis is log-scaled, DeepSeek is much cheaper than visually implied (mousing over the raw values, it's 1/4th the cost of Luna).
125.
by minimaxir 2mo ago
It's always fun when Max reasoning is cheaper than High reasoning.
126.
by minimaxir 2mo ago
There are a lot of Monaco billboards up now which made me laugh.
127.
by minimaxir 2mo ago
It's more of a tragedy of the commons. There are so many AI billboards in SF that I legitimately can't differentiate any of the AI startups anymore, and that's even if they say what their product actually is .
128.
by minimaxir 2mo ago
OpenAI tried auto-routing with the initial GPT-5 release and it was immediately clear why that was a bad idea.
129.
by minimaxir 2mo ago
Which was in itself a do-over because Opus 4.7 received a lot of bad press on suspicion of being a regression from 4.6.
130.
by minimaxir 2mo ago
Your entire commenting history is about how much you dislike LLMs (with the high likelihood that you would not accept any evidence about agentic productivity), which does not lead to good discourse on Hacker News.
131.
by minimaxir 2mo ago
Which is in itself a bit weird as mid models nowadays are a golden mean fallacy. Terra is much less popular than both Luna (cost-sensitive) and Sol (performance-sensitive). Claude Sonnet is a weird exception to the mid models because Anthro
132.
by minimaxir 2mo ago
Muse Spark 1.1 was released July 16th, less than a month ago. A new version release this soon (particularly after Kimi K3's release drastically overshadowed it) is a bit sus and it appears that Meta is trying a first launch do-over.
133.
by minimaxir 2mo ago
This repo is all "stolen code regurgitated by an LLM" so that is an impossible task and therefore makes the request performative.
134.
by minimaxir 2mo ago
The README ( https://github.com/adamtwiss/coda#license ) addresses concerns about copyright in good faith and better than most agentic-driven repos: > After our initial 0.9.0 pre-release (July 2026), community feedbac
135.
by minimaxir 2mo ago
I'm particularly annoyed with the complaint in the linked thread about the agentic-developed chess engine. > Moreover, you are giving absolutely nothing back to the community, which is central to open-source engine dev. You are on
136.
Sandboxing(zed.dev)
5 points by minimaxir 2mo ago|4 comments
137.
by minimaxir 2mo ago
The rankings are for prompt adherence, not subjective quality.
138.
by minimaxir 2mo ago
Minimax just released the weights for a video model ( https://huggingface.co/MiniMaxAI/MiniMax-H3 ) that apparently only has limited safeguards.
139.
by minimaxir 2mo ago
The scores are pairwise ELO rankings, not simple metric scores.
140.
by minimaxir 2mo ago
Everything is less good than gpt-2-image and I suspect that will be the case for awhile, until potentially Nano Banana Pro 2. However, cost is significantly lower in this case. A Pro image here is $0.04, a gpt-2-image high is $0.21 and lo
141.
by minimaxir 2mo ago
Also available on OpenRouter (whose image output support is getting better): https://openrouter.ai/qwen/qwen-image-3-pro
142.
by minimaxir 2mo ago
The model was added to OpenRouter yesterday.
143.
by minimaxir 2mo ago
The drama around LLM use in game decompiling and other related hacking is very funny. I've seen "well maybe you should learn C++ instead of being lazy and vibecoding" a nonneligible number of times.
144.
by minimaxir 2mo ago
> Harassment is not allowed . You may not harass people for using an LLM, regardless of whether or not their use is banned by the policy. You must follow the Code of Conduct at all times when interacting with the Rust project. This poli
145.
by minimaxir 2mo ago
Location: San Francisco / Bay Area Remote: No Willing to relocate: No Technologies: pretty much everything related to data science/machine learning/AI Résumé/CV: https://minimaxir.com/ (resume available
146.
by minimaxir 3mo ago
Unfortunately people are saying they are fun for the purposes of ragebait that goes viral, which is half the reason people intentionally post AI slop on social media.
147.
by minimaxir 3mo ago
> so people have less of an incentive to be honest, and tons of incentives to hide their use Case in point, the recent Hank Green incident. That was an own-goal of being too honest about LLM use and I suspect the industry learning will b
148.
by minimaxir 3mo ago
> 1. You believe in ascribing credit where credit is due. Using an LLM feels like cheating so disclosing it makes you feel less like an imposter for having produced something amazing. Cheating isn't the right word here. Crediting LL
149.
by minimaxir 3mo ago
...that links to a Medium blog post, this submission is not a Medium blog post. I can confidently say I do not have an encyclopedic knowledge of blog posts and who wrote them I may or may not have read over a year ago. I made that comment o
150.
by minimaxir 3mo ago
I have never read your blog before this submission. Technical blog posts don't need to follow AP Style but this post reads like LinkedIn broetry.

More ›