Thinky releases open-source reasoning model optimized for post-training

@Awesome_O_AI Itd a game changer for most usecases! Doing the same tasks in half as much time with half the resources is always a huge unlock

1:32 PM · Jul 15, 2026

Congrats to the @thinkymachines on an amazing model!! Thank you @johnschulman2 and @miramurati for your contributions to open science!

12:11 PM · Jul 15, 2026

Combined views

151

2 posts, first seen 9h ago

15 reposts

Sentiment

Positive100%0%Negative

Positive Read

Users expressed excitement over Thinky's open source model optimized for agents and customization, praising it as a game changer that halves task time and resources while advancing open science.

Based on 4 visible X reactions from 3 accounts; directional sample.

Digg Deeper

Ask a question below.

Published answers will appear here.

Commentary on X

SG

Simon Guo@simonguozirui

@MichaelElabd thank you so much for testing with us and always providing such detailed feedback! excited for you all to train more models with it and @tinkerapi!

11:46 AM · Jul 15, 2026

@Awesome_O_AI Itd a game changer for most usecases! Doing the same tasks in half as much time with half the resources is always a huge unlock

1:32 PM · Jul 15, 2026

Congrats to the @thinkymachines on an amazing model!! Thank you @johnschulman2 and @miramurati for your contributions to open science!

12:11 PM · Jul 15, 2026