Is that no longer the case, or am I misunderstanding the operational costs displayed?

Opus: 521k input tokens; 12k out

Grok: 443k input tokens; 57k out

Gemini: 677k input tokens; 7k out

OAI: 543k input tokens; 17k out

Gemini appears to use by far the least amount of reasoning tokens, assuming they're included in the output counts.

How many rooms in your home though? These are crucial details.

I’m obviously not answering that without a long painful discussion of what constitutes a room.

I have a feeling you are using SOTA models at work and aren't used to just how cheap the non-Anthropic/Google/OAI options are these days. GLM's coding subscription is like $6/month if you buy a full year.

Reading the parent comment, I assume their use of 'freedom' more closely aligns with being undisturbed by a government.

It's a very common usage in America, focusing on "Freedom from X" rather than "Freedom to do Y", the latter of which often needs some sort of societal protection, most often provided by said government.

I use monarch and I've been happy enough with it. Would probably consider self-hosting with actual in the future, but I wanted an easy on-ramp for myself to actually get in the habit of budgeting.

Hi, that's not why people go to Delaware for those kind of purchases. It's the lack of tax.

The actual MSRP from PA wine and liquor stores is very competitive, since it's one of the largest single buyers of alcohol. Selection could be better though.

Hard agree. I may cross state lines for large purchases like laptops, but the liquor prices aren’t so bad that I’d drive >30min for better deals.

I don't think this is an accurate comparison. Working on open source software means you are contributing to that software, which yes may be used by for profit companies. This is more analogous to contributing to Wikipedia, which is then used by for profit companies like Grok, than it is contributing to Grok products directly, which cannot be leveraged by other tools in this ecosystem (afaik).

Small counterpoint but there are also 2 new players putting out SOTA open source models (Moonshots Kimi and zhipus GLM) so we're still seeing the same number of models overall, just via newer entrants.

> GLM-4.6 is great value but still not solid enough for tool calls, not that fast, etc. so if you can afford something more reliable I'd go for that, but encouraging.

Funny you should say that, because while it is a large model the GLM 4.5 is at the top of Berkley's Function Calling Leaderboard [0] and has one of the lowest costs. Can't comment on speed compared to those smaller models, but the Air version of 4.5 is similarly highly-ranked.

[0]https://gorilla.cs.berkeley.edu/leaderboard.html

Gorilla is a great resource and it isn't unreasonable to suspect Z.AI has it in their data sets. I'd suspect most other frontier labs as well (pure speculation, but why not use it as a resource).

Problem is, while Gorilla was an amazing resource back in 2023 and continues to be a great dataset to lean on, but most ways we use LLMs in multi step tasks have since evolved greatly, not just with structured JSON (which GorillaOpenFunctionsV2, v4 eval does multi too), but more with the scaffolding around models (Claude Code vs Codex vs OpenCode, etc.). Likely why good performance with Gorilla doesn't necessarily map onto multiple step workloads with day-to-day tooling, which I tend to go for and reason why, despite there being FOSS options already, most labs either built their own coding assistant tooling (and most open source that too) or feel the need to fork others (Qwen with Geminis repo).

Purely speculative, but GLM-4.6 I evaluated using the same tasks as other models via Claude Code with their endpoint as that is what they advertise as the official way to use the model, same reason I use e.g. Codex for GPT-5. More focused on results in the best case, over e.g. using opencode for all models to give a more level playing field.

I didn't really follow the premise of the original video. Can you find identical food at various locations throughout the country? Yes. Are you required to eat this same food? No!

You could likely order the things that local restaurant/pub does make in house, rather than go for the variety of items on the menu that are simply there to appease customers' fancies.

Or you could go to a proper restaurant that makes the majority of their dishes in house. You have this choice.

If Sysco and these other companies didn't offer pre-made jalepeno poppers and the like, they simply wouldn't be on the menu in the first place.

That's unrealistic. The problem is most restaurants in America exist in rural areas in commercial food deserts and are almost entirely dependent upon mega broadline suppliers that have merged into de facto oligopolies. Furthermore, by offering a wealth of finished items, in addition to ingredients and supplies, it's seductive to restaurants use these boring, lazy, pre-made offerings rather than cooking from scratch like respectable restaurants that do the work and pay the cost of sourcing quality ingredients, taking the time to do the preparation, and cooking good food.

What's a good way to tell which things are made in house and which are mass market?

And I think Sysco acquiring other local suppliers does suggest that more menu items might be from them than were in the last at least

Ask.

how are customers going to tell in-house dishes from sysco premade dishes just by looking at a menu?