I found Ruby LLM to be surprisingly good - in terms of usability it's close to Vercel's AI framework.
It tries to strike a balance between working out of the box and being flexible... which has its challenges, still nice overall.
One big real-life pain I experienced is that caches don't always work, e.g. for xAI, since it only supports completions API and thought signatures are returned wrong.
RubyLLM is very easy to use. Made extensive use of it for a project last year. Drawbacks are it was difficult to instrument for true trace observability and it has a pattern where retries will delete the underlying models so the history you see is clean but not necessarily great for seeing exactly what the sequence of API calls was.
It is quite nice, but not as nice as you'd want. You still have to set platform specifics when running completions when you want to tune things like temperature, effort, max tokens, etc.
Hi! Valid challenge, I am probably misremembering. We were playing with various 'one-interface to all providers' solutions and I might have mixed up RubyLLM there. Sorry for that.
I will have a deep dive into which things I felt we needed to adapt per provider.
I didn't mean to imply that you have to solve all of our wants of course.
One thing we did do was monkey-patch the spot where tool_calls are performed by RubyLLM. We had our own mechanism for that and were able to skip RubyLLM's and still extract the tool calls and run them through our own tool harness. That all worked beautifully. I don't know if that type of stuff is something you want PRs on or that you want to keep steering towards the route that does everything within RubyLLM classes. Happy to contribute some of that.
Why would anyone still build in dynamically typed languages in 2026? Why relinquish the crystal clear signals that static typing is able to provide to the LLM?
I found Ruby LLM to be surprisingly good - in terms of usability it's close to Vercel's AI framework.
It tries to strike a balance between working out of the box and being flexible... which has its challenges, still nice overall.
One big real-life pain I experienced is that caches don't always work, e.g. for xAI, since it only supports completions API and thought signatures are returned wrong.
Thank you!
Responses API is now implemented and it's coming in RubyLLM 2.0
https://github.com/crmne/ruby_llm/blob/main/lib/ruby_llm/pro...
Do you have any details published around 2.0? Would love to learn more.
Not yet. I'll do a series of blog posts and tweets in the next weeks.
RubyLLM is very easy to use. Made extensive use of it for a project last year. Drawbacks are it was difficult to instrument for true trace observability and it has a pattern where retries will delete the underlying models so the history you see is clean but not necessarily great for seeing exactly what the sequence of API calls was.
Glad you like it.
Rails-style instrumentation landed in 1.16.0.
https://rubyllm.com/instrumentation/
I have an open source gem called Raix that builds on top of RubyLLM's abstractions and is quite popular. https://github.com/OlympiaAI/raix
thank you for bringing ruby into AI community and your open-source work. Great language must be explored and get more attention :)
Thank you!
I love how MINASWAN Hacker News is when talking about Ruby!
I built a similar Ruby based agent development kit that has a different focus and feature set:
https://github.com/tweibley/legate
It is quite nice, but not as nice as you'd want. You still have to set platform specifics when running completions when you want to tune things like temperature, effort, max tokens, etc.
RubyLLM author here.
I'm not sure where you got that.
`chat.with_temperature(0.2)`
https://rubyllm.com/chat/#controlling-response-behavior
`chat.with_thinking(effort: :high, budget: 8000)`
https://rubyllm.com/thinking/#controlling-extended-thinking
Max tokens is the only one of your list that require provider specific params:
https://rubyllm.com/chat/#provider-specific-parameters
I'm one guy doing it for free. Happy to see your contribution!
Hi! Valid challenge, I am probably misremembering. We were playing with various 'one-interface to all providers' solutions and I might have mixed up RubyLLM there. Sorry for that.
I will have a deep dive into which things I felt we needed to adapt per provider.
I didn't mean to imply that you have to solve all of our wants of course.
One thing we did do was monkey-patch the spot where tool_calls are performed by RubyLLM. We had our own mechanism for that and were able to skip RubyLLM's and still extract the tool calls and run them through our own tool harness. That all worked beautifully. I don't know if that type of stuff is something you want PRs on or that you want to keep steering towards the route that does everything within RubyLLM classes. Happy to contribute some of that.
Interesting! What were you guys trying to achieve by running them in your own tool harness?
And thank you! It is absolutely awesome and a true joy to work with.
I have created an open source chatgpt clone with rubyllm, check it out here: https://www.railschat.org/
Thank you
In case you're using PHP or Node.js, we've made a similar toolkit free and open source on github: https://github.com/Qbix/AI/tree/main/classes/AI
Why would anyone still build in dynamically typed languages in 2026? Why relinquish the crystal clear signals that static typing is able to provide to the LLM?
Well, LLMs have an obscene amount of context built into their weights about Ruby on Rails, and can work within it extremely quickly.
This is not a tool for using LLMs to write Ruby code.