/AI2h ago

Daniel Han, Unsloth creator, launches OpenEnv to standardize reinforcement learning interfaces with backing from Hugging Face, Meta-PyTorch, and NVIDIA

It decouples AI models from tightly bound RL training environments

22187276027.5K
#17Thomas Wolf
clem 🤗
will brown
Daniel Han
Charles 🎉 Frye
Ben Burtenshaw
Original postclem 🤗#67
Ben Burtenshaw@ben_burtenshaw

So excited to be opening up OpenEnv to the whole community. It will now be owned by @huggingface , Meta-PyTorch, @reflection_ai , @UnslothAI , @modal, @PrimeIntellect , @NVIDIAAI , @mercor_ai , and @fleet_ai .

the reason is: frontier labs train the model and the harness together, so the model is fitted to its harness. that coupling is a chunk of why claude code and codex feel so good.

open source can't do that. you bring whatever harness, whatever model, whatever env, whatever trainer. which is the whole point of open source and also the problem for training.

openenv is the socket in between all of this.

in short: it's a protocol layer, not a reward framework. it does not have opinions about your rewards or your training loop. those live in the libs that are actually good at them.

read more in the blog post. it's early, come break it.

7:27 AM · Jun 8, 2026 · 22.5K Views
Sentiment

Users are enthusiastic about OpenEnv opening to community ownership with Hugging Face and NVIDIA backing because the collaborations separate models from training harnesses and solve open-source RL bottlenecks while speeding development.

Pos
100.0%
Neg
0.0%
5 comments with sentiment.
Cluster Engagement
Posts from X
Most Activity
Most Activity
VIEWS5.2KBOOKMARKS10LIKES53
will brown@willccbb

open environments for everyone 🫡

Ben Burtenshaw@ben_burtenshaw

So excited to be opening up OpenEnv to the whole community. It will now be owned by @huggingface , Meta-PyTorch, @reflection_ai , @UnslothAI , @modal, @PrimeIntellect , @NVIDIAAI , @mercor_ai , and @fleet_ai .

the reason is: frontier labs train the model and the harness together, so the model is fitted to its harness. that coupling is a chunk of why claude code and codex feel so good.

open source can't do that. you bring whatever harness, whatever model, whatever env, whatever trainer. which is the whole point of open source and also the problem for training.

openenv is the socket in between all of this.

in short: it's a protocol layer, not a reward framework. it does not have opinions about your rewards or your training loop. those live in the libs that are actually good at them.

read more in the blog post. it's early, come break it.

1hViews 5.2KLikes 53Bookmarks 10
RETWEETS4
Daniel Han@danielhanchen

Super excited Unsloth will be collabing on OpenEnv to make RL even more accessible to everyone!

OpenEnv is the interface between the harness (eg Codex, Claude Code), the RL environment, and the trainer! The goal is to make RL truly plug n play for everyone!

Ben Burtenshaw@ben_burtenshaw

So excited to be opening up OpenEnv to the whole community. It will now be owned by @huggingface , Meta-PyTorch, @reflection_ai , @UnslothAI , @modal, @PrimeIntellect , @NVIDIAAI , @mercor_ai , and @fleet_ai .

the reason is: frontier labs train the model and the harness together, so the model is fitted to its harness. that coupling is a chunk of why claude code and codex feel so good.

open source can't do that. you bring whatever harness, whatever model, whatever env, whatever trainer. which is the whole point of open source and also the problem for training.

openenv is the socket in between all of this.

in short: it's a protocol layer, not a reward framework. it does not have opinions about your rewards or your training loop. those live in the libs that are actually good at them.

read more in the blog post. it's early, come break it.

2hViews 2.6KLikes 34Bookmarks 5
REPLIES5
Thomas Wolf@Thom_Wolf

"Starting today, OpenEnv will be coordinated by a committee that so far includes Meta-PyTorch, Reflection, Unsloth, Modal, Prime Intellect, Nvidia, Mercor, Fleet AI, and Hugging Face."

Excited to keep growing the collaboration behind the open agentic RL stack!

Read more at https://huggingface.co/blog/openenv-agentic-rl

Ben Burtenshaw@ben_burtenshaw

So excited to be opening up OpenEnv to the whole community. It will now be owned by @huggingface , Meta-PyTorch, @reflection_ai , @UnslothAI , @modal, @PrimeIntellect , @NVIDIAAI , @mercor_ai , and @fleet_ai .

the reason is: frontier labs train the model and the harness together, so the model is fitted to its harness. that coupling is a chunk of why claude code and codex feel so good.

open source can't do that. you bring whatever harness, whatever model, whatever env, whatever trainer. which is the whole point of open source and also the problem for training.

openenv is the socket in between all of this.

in short: it's a protocol layer, not a reward framework. it does not have opinions about your rewards or your training loop. those live in the libs that are actually good at them.

read more in the blog post. it's early, come break it.

43mViews 2.3KLikes 15Bookmarks 6
Vidit Ostwal@ViditOstwal

@ben_burtenshaw @huggingface @reflection_ai @UnslothAI @modal @PrimeIntellect @NVIDIAAI @mercor_ai @fleet_ai Adding the blog post here, I got a bit confused where to find this: https://huggingface.co/blog/openenv-agentic-rl

2hViews 191Likes 4Bookmarks 1
The AI Therapist@TheAIShrink

@ben_burtenshaw @huggingface @reflection_ai @UnslothAI @modal @PrimeIntellect @NVIDIAAI @mercor_ai @fleet_ai Eight organizations own it now. Avoids one company controlling the standard. Introduces committee shipping. Which is how you get every good idea AND every compromise.

2hViews 99Likes 1
每日ai@DailyAI_1

@ben_burtenshaw @ClementDelangue @huggingface @reflection_ai @UnslothAI @modal @PrimeIntellect @NVIDIAAI @mercor_ai @fleet_ai 感觉挺有趣的？

51mViews 48Likes 1
Piyush@CatAstro_Piyush

@danielhanchen 🔥

2hViews 38Likes 1
Sentio@Sentio_xbt

@willccbb Open tends to win eventually

1hViews 11Likes 1
haro@harobuilds

@Thom_Wolf separating the model from the harness is the right call. whoever owns the training environment owns the benchmark and that's where the real lock-in lives

43mViews 4Likes 1
Rugbist@rugbist_

@danielhanchen plug n play RL sounds like a dream until something silently breaks at step 2

curious what harnesses are confirmed day one

2hViews 8
Paco@Pacoxbt

@Thom_Wolf Big names coming together for open RL

this kind of collaboration usually speeds up development and keeps things open interesting to see meta and nvidia involved alongside open source focused groups

38mViews 3
Blissy@BlissyOnX

@Thom_Wolf committee reads like a whos who of things that are definitely shipping in 2025

missing a brokered episode though

41mViews 2
Invincible@InvincibleEdge

@Thom_Wolf this actually solves the open-source training bottleneck nicely

huggingface + pytorch parent + nvidia is the real power move here

37mViews 1
Rugbist@rugbist_

@Thom_Wolf committee packed with heavy hitters, hope the actual community still gets a real voice in there

41mViews 1