Introducing October, the world's most advanced agent orchestrator Claude? Codex? Grok? Cursor? Which model? Which harness? How much quota is left? You shouldn't have to care Give October a task. It builds the team, connects the agents with native agent agent comms, and orchestrates the work based on the... task, your taste, budget and remaining usage (or build the team yourself) And it's multiplayer: your team's agents can work with yours, live on the same canvasshow more

just tried it, actually so freaking awesome. I love the Kanban.

Glad to hear! Would you have time to jump on a call? Would love more feedback!

bet bet

This is amazing bro. Love that @october_ai takes the whole model and quota decision out of the way and lets you start with the work itself. The way it builds a team, gets the agents to cross check each other and turns the workflow into something reusable feels like a huge step forward. Congratulations on the launch bro, excited to see where you take this.

@october_ai Yes! The whole point of a great orchestrator is that it takes all the hard and boring parts out of your way!

@october_ai Exactly. LFG🚀

@october_ai 🙌🙌

@Ozacle23 Insanee, @harshsaver could you check dms as well!!

@Ozacle23 Thanks Shivank! Yes checking

@Ozacle23 Thanks ser

interesting UI concept, how does claude talk to codex? what do you do to make that work? can i use deepseek or kimi?

we have built a protocol called October bus! It works with 30+ harnesses, and it let’s claude code / codex etc send messages to each other!

Excited to try this out!

Please do! This is especially an amazing layout for working on hardware and electronics project! You can see every part of the flow and create blueprints + code!

i used october this past 2 months, claude orchestrating open source models has saved me a lot of money

i am so glad to hear that! please let me know if you have any feedback!

This is super cool

thank you so much! 🙏🙏

Agent army 💀

On October you can do agent army duels haha

def the most advanced agent orchestrator! super cool.

yessir! the harness of the harnesses!

This is useful, given we’ve reached a point where it’s impossible to keep track of & test every new launched model

+ mixture of models and harnesses has been giving great results! rather than Fable one shotting something or Astra one shotting something - Fable + Astra one shotting something together catches more issues and edge cases also you can save money! get fable to plan and grok to execute and so on

Amazing things, what a pleasure to have seen this develop from a great idea to such a powerful platform🔥🎃

❤️❤️ so glad to have you on October from day 1!

As said before and I’ll say it again, to great times ahead with you guys! Wether that be behind the screen or the camera, we’ll have enough to do together on this insanity😏

🫡🫡❤️

i downloaded october a few days ago and really like the product.

thanks! keep using and sharing feedback!

this is super cool!!!!

This is soo cool!

🙌🙌 orchestrate your work with October!

Goated product

🙌🙌🔥🔥🙏

great work, harsh. congrats on the launch 🙌

thank you Mahesh! 🫡🙌 Posting about the Supermemory integration soon!

Let’s gooooo

Meetstream agents on October? 👀

Let's do it!

🙌🙌

wohoooo! all the best ❤️

thank you Ojasvika! 🙌

Don’t miss out on this peops, this is the frontier in multi agent work 🙌

thank you Reha! and it works with all your current subscriptions!

Sick! agent lego ftw

haha I love it - agent lego!

launching October in September 🔥

You are not gonna believe it - I need to figure out when to launch November and December - two of October’s open source projects 😂

Damn, this update is looking sweet!!

Thanks Justin! Keep sharing feedback!

Crazzzyyyyyyyyyy🫡🫡🫡🫡🫡🫡

🙌🙌

Awesome!

thanks sir!

i have 8 agent accounts T_T i will use anything to save me tokens + money checking this out

yep, with October you don’t have to keep feeding context to all agents! You can just let them share the most minimum required context

@VedSheth21 OCTOBER FTW🚀🚀🚀🚀

@VedSheth21 lessgooo 🤧

How does October prevent agents from duplicating work or declaring success before the result is verified?

we are testing all the edge cases of this! right now all agents have an inbox and a shared todo list, so every agent can check which task any other agent has picked - and then it doesn’t pick the same task

this looks way better than before, nice work

thanks! keep sharing feedback!

Insane work, congrats 🙌

thanks sir 🫡 what do you think @julianweisser 👀

Can you swap model mid-project or is it set once

you can swap any time - just like you would in a normal terminal!

Awesome 🔥

FINALLYYY

finally finally finally ❤️🙌

cool stuff

thank you! 🫡

Have seen October evolve first hand and it's crazyyy!

yess! it’s now so much better than before! we ship daily 💪

The real product is task-level orchestration: agents chosen and coordinated around constraints like budget, quota, and user preferences. A shared canvas becomes valuable when it preserves context and makes handoffs inspectable.

and hand offs between apps too! Not just agents

works with claude sub i assume? :D

yep! 30+ harnesses directly work with their subscriptions!