And let's not forget that they regularly break things. For example they had merged a pull request that was supposed to fix issues related to the openrouter/free models but in the process broke openrouter for all models except the free ones.

The fix was deployed 2 weeks later (!!) to main (but you could downgrade of course).

Or that other time they broke model selection if you had selected "this key can used all models of their team" than the only model in the auto-selection for harnesses was an invalid "all-team-models" entry. Fixes this one in 1 week though.

all of them on the :latest docker tag btw

The best solution to this for me is to self host litellm or a different router and use model aliases. For example I have a model called "coding" and when a new good model comes out I just switch the backend without needing to change the alias or the key in my projects (opencode, etc).

I have a few of them even a smart router called "agents" which will use local models but if it thinks the request might require higher reasoning it's routing to a different model

Yeah, I also built my own "router" for this:

    if (process.env.LOCAL_MODEL {
      http('localhost:3000/v1/completions')...
    } else {
      http('api.openrouter.ai/v1/completions')...
    }

Brilliant! I'd like to invest $10M. Who should I make the check out to?

investors REALLY want the closed models to stay frontier forevery. mmw the future of AI will be local and offline

and pretty sure they're gonna drop crypto currencies as payment on the platform too

I doubt it. One of the founders was on a podcast recently and they were singing stablecoin praises.

Why? How do these relate?

Metronome has been experimenting with real-time payment on usage based billing for tokens. That technology leverages micro payments using crypto. It's cool but all it's doing is moving the accumulated balance from a provider onto your crypto wallet.

and memory bandwidth

Do you mean 3.6 27b? Because qwen 3.7 didn't have an open weight version

won't forgive them either for making it harder to filter out ads

it does! I'm always amazed when I use DSV4 flash for coding or server checks and after an hour of working with it my (pure api call) balance is about 30 cents

Sorry the title should have been "You should self-host your mail server" but it seems HN removes the "You should" in the beginning of a title.

I wanted to start this discussion because I think we are prime for more self hosted mail servers in 2026 and in the post I explain how to use a locally hosted LLM to nudge the spam detector rspamd to whether or not an incoming mail is spam.

Also in times where the EU can order your mail providers to scan for unwanted content [1] (currently limited to CSAM, later probably more) it does feel better to have your data at your hand.

[1] https://en.philenews.com/international/privacy-to-end-in-the...

it would help if they used strict types and modern standards but as you say the wordpress codebase is beyond dated and held together with duckt tape