[rbren]

I’m biased [0], but I think we should be scripting around LLM-agnostic open source agents. This technology is changing software development at its foundations—-we need to ensure we continue to control how we work.

[0] https://github.com/all-hands-ai/openhands

[robotbikes]

This looks like a good resource. There are some pretty powerful models that will run on a Nvidia 4090 w/ 24gb of RAM. Devstral and Queen 3. Ollama makes it simple to run them on your own hardware, but the cost of the GPU is a significant investment. But if you are paying $250 a month for a proprietary tool it would pay for itself pretty quickly.

[NitpickLawyer]

> There are some pretty powerful models that will run on a Nvidia 4090 w/ 24gb of RAM. Devstral and Queen 3.

I'd caution against using devstral on a 24 gb vram budget. Heavy quantisation (the only way to make it fit into 24gb) will affect it a lot. Lots of reports on locallama about subpar results, especially from kv cache quant.

We've had good experiences with running it fp8 and full cache, but going lower than that will impact the quality a lot.

[seanmcdirmid]

A Max M3 with 64 GB works well for a wider range of models although it fairs worse on stable diffusion jobs. Plus you can get it as a laptop.

[handfuloflight]

But what do we do if the closed models are just better?

[rkangel]

The agents are separate from the models. Claude Code only allows you to use Claude, but Aider allows you to use any model.

[handfuloflight]

How does that solve the problem of closed models being better than open models?

[hn8726]

There is no problem. OP said we should be using open _agents_, not open _models_. You can use an open agent with any model, open or closed, while using something like Claude Code locks you in to one model vendor

[handfuloflight]

I know what OP said and I asked a question in turn.

[davidmurdoch]

Wait?

[handfuloflight]

And get superseded by competitors willing to spend on those models?

[bluefirebrand]

Steal from them shamelessly, the same way they stole from everyone else?

[hsuduebc2]

You are onto something.

[datameta]

Seems ethically sound to me.

[dghlsakjg]

Isn't abusing the OpenAI terms of service part of how Deepseek did training?

[ProofHouse]

This 10000%