Pickaxe.co

This is the platform I use to house Kai (LLM). I’ve looked at and tried other platforms, but in my opinion, this is the best place to start.

I consider myself a committed hobbyist, not someone launching a full-blown business, and Pickaxe has been the perfect place for me to build, experiment, and keep everything in one spot. If you are starting a business, I still think it’s a great place to begin—it’s straightforward, flexible, and welcoming.

You can also sign up for a free account and get a feel for what it’s like to design a custom landing spot for your bot. Trust me . . . this part is fun.

I’ll also say this: the people at Pickaxe have been genuinely helpful. I’m no expert in the world of AI, and I’ve had plenty of questions. Their support has been solid, and there’s also a community for asking general questions, which is nice when you’re still figuring things out. I’m still a little shy about jumping in too deeply, but it’s nice to know it’s there.

If you decide to sign up for a paid plan, you can get a 10% discount by using the link here.

ElevenLabs. io

If you want to give your bot a voice, I recommend ElevenLabs. The voices are genuinely impressive, and once everything is set up correctly, the result can feel a little magical.

Getting an API key and Voice ID from ElevenLabs is straightforward enough, and you can connect ElevenLabs to Pickaxe through the Actions Library.

A quick note before you fall in love with the idea: if you’re using Pickaxe, voice is not available on the free plan. You’ll need at least the Gold Plan to use this voice Action, and the Gold Plan currently allows up to three Actions per bot.

Now for the part that is less glamorous: setting it up was not plug-and-play for me.

As of this writing, Pickaxe supports Eleven Multilingual v1, so that is the version you need to choose for the connection to work properly. I also had to add extra instructions at the end of the bot’s Role command under Configure to make the voice action behave correctly.

The instruction I used was:

“In generating the full response, call the ElevenLabs action and pass the entire response text into it.

In your reply, include:
The full written response.
The audio file returned by the ElevenLabs action.”

Once I got all of that sorted out, it worked beautifully. But I want to be honest: the setup took patience, troubleshooting, and more fiddling than I expected. So yes, I recommend ElevenLabs for the voice quality—but not because the setup is effortless. I’ll be writing a fuller post about this setup, because it took more trial and error than I expected.

If you decide to try it, you can use the link here to get 50% off your first month.

Stay Tuned:

As I keep testing tools, chasing glitches, and discovering new favorites, I’ll post the updates here.