Benchmarking AI Gateways: GoModel vs. LiteLLM vs. Portkey vs. Bifrost
enterpilot.ioGoModel author is here.
This is my attempt to honestly compare GoModel vs LiteLLM vs Portkey vs Bifrost AI gateways.
Very interesting, I really wish that Python would stop being used for core things like gateways, just way way too inefficient and too much overhead. Great to see other options and the strengths of non-python stacks being highlighted, looking forward to trying out GoModel.
There seems to be movement within the Python/TypeScript ecosystem to replace implementations with Rust and continue distribution within said ecosystem.
I'm also looking to try out GoModel this weekend, met the creator in the OpenCode discord this last week. Looking to pair that with https://github.com/kubeopencode/kubeopencode
Thank you! Feel free to reach out to me over Discord in the case of any feedback or feature request.
Appreciate the reproduction, perhaps we can expand it to cover more providers / local and build out a larger perspective.
Thoughts on breaking out the benchmarking code into a separate repo?
Currently, I want to keep them close to the documentation, where I link to them directly. The fewer repositories I have, the easier it is to maintain them.
lol, are we headed towards monAIrepos?
I mentally treat it as part of the documentation right now.
Yeah, I hear you on that, I've been using what I call "mega-repos" where all the things for a single project live in the same repo. I retract my prior request upon reflection :]