DeepReinforce releases open-source Ornith-1.0 coding models, claiming its 397B MoE beats Claude Opus 4.8 on Terminal-Bench

VIEWS150.8KLIKES1.2KREPLIES77

Chubby♨️@kimmonismus

This looks to good to be true.

A 397B open source model on par or even outperforming Claude Opus 4.8?

I need to check it out.

Ornith@ornith_

Aloha! 🌺 Meet Ornith-1.0, a family of open-source LLMs specialized for agentic coding.

Ornith-1.0 spans the full parameter sizes including 9B Dense, 31B Dense, 35B MoE, and 397B MoE. It achieves state-of-the-art performance among open-source models of comparable size on coding benchmarks including: ✅Terminal-Bench 2.1(77.5) ✅SWE-Bench(82.4 on verified, 62.2 on pro, 78.9 on Multilingual) ✅NL2Repo(48.2) ✅SWE Atlas(41.2 on QnA, 42.6 RF, 39.1 TW) ✅ClawEval(77.1)

Post-trained on top of gemma4 and qwen3.5, Ornith-1.0 employs a novel self-improving training strategy in which reinforcement learning is used to generate not only solution rollouts, but also the task-specific scaffolds that drive those rollouts. By jointly optimizing the scaffold and the resulting solution, the model generate higher-quality solutions in agentic coding.😎

All models are released under the MIT license, enabling full commercial and research use.

📖Tech Blog: http://deep-reinforce.com/ornith_1_0.html 🤗Huggingface: http://huggingface.co/collections/deepreinforce-ai/ornith-10

21h150.8K1.2K366

BOOKMARKS601

Brian Roemmele@BrianRoemmele

We are testing US open source Ornith 1.0 and it is absolutely amazing.

The last nail in the Anthropic Coffin!

Good work folks!

Ornith@ornith_

Aloha! 🌺 Meet Ornith-1.0, a family of open-source LLMs specialized for agentic coding.

Ornith-1.0 spans the full parameter sizes including 9B Dense, 31B Dense, 35B MoE, and 397B MoE. It achieves state-of-the-art performance among open-source models of comparable size on coding benchmarks including: ✅Terminal-Bench 2.1(77.5) ✅SWE-Bench(82.4 on verified, 62.2 on pro, 78.9 on Multilingual) ✅NL2Repo(48.2) ✅SWE Atlas(41.2 on QnA, 42.6 RF, 39.1 TW) ✅ClawEval(77.1)

Post-trained on top of gemma4 and qwen3.5, Ornith-1.0 employs a novel self-improving training strategy in which reinforcement learning is used to generate not only solution rollouts, but also the task-specific scaffolds that drive those rollouts. By jointly optimizing the scaffold and the resulting solution, the model generate higher-quality solutions in agentic coding.😎

All models are released under the MIT license, enabling full commercial and research use.

📖Tech Blog: http://deep-reinforce.com/ornith_1_0.html 🤗Huggingface: http://huggingface.co/collections/deepreinforce-ai/ornith-10

20h141.3K1.2K601

RETWEETS498

Ornith@ornith_

Aloha! 🌺 Meet Ornith-1.0, a family of open-source LLMs specialized for agentic coding.

Ornith-1.0 spans the full parameter sizes including 9B Dense, 31B Dense, 35B MoE, and 397B MoE. It achieves state-of-the-art performance among open-source models of comparable size on coding benchmarks including: ✅Terminal-Bench 2.1(77.5) ✅SWE-Bench(82.4 on verified, 62.2 on pro, 78.9 on Multilingual) ✅NL2Repo(48.2) ✅SWE Atlas(41.2 on QnA, 42.6 RF, 39.1 TW) ✅ClawEval(77.1)

Post-trained on top of gemma4 and qwen3.5, Ornith-1.0 employs a novel self-improving training strategy in which reinforcement learning is used to generate not only solution rollouts, but also the task-specific scaffolds that drive those rollouts. By jointly optimizing the scaffold and the resulting solution, the model generate higher-quality solutions in agentic coding.😎

All models are released under the MIT license, enabling full commercial and research use.

📖Tech Blog: http://deep-reinforce.com/ornith_1_0.html 🤗Huggingface: http://huggingface.co/collections/deepreinforce-ai/ornith-10

1d4.8M5.9K5.7K

Teortaxes▶️ (DeepSeek 推特🐋铁粉 2023 – ∞)@teortaxesTex

almost all engagement with this seems synthetic I'd be interested in serious people like @_xjdr giving it a spin, until then I reserve my skepticism that Qwen 3.5 base can approach or surpass GLM 5.2 just like that

Ornith@ornith_

Aloha! 🌺 Meet Ornith-1.0, a family of open-source LLMs specialized for agentic coding.

Ornith-1.0 spans the full parameter sizes including 9B Dense, 31B Dense, 35B MoE, and 397B MoE. It achieves state-of-the-art performance among open-source models of comparable size on coding benchmarks including: ✅Terminal-Bench 2.1(77.5) ✅SWE-Bench(82.4 on verified, 62.2 on pro, 78.9 on Multilingual) ✅NL2Repo(48.2) ✅SWE Atlas(41.2 on QnA, 42.6 RF, 39.1 TW) ✅ClawEval(77.1)

Post-trained on top of gemma4 and qwen3.5, Ornith-1.0 employs a novel self-improving training strategy in which reinforcement learning is used to generate not only solution rollouts, but also the task-specific scaffolds that drive those rollouts. By jointly optimizing the scaffold and the resulting solution, the model generate higher-quality solutions in agentic coding.😎

All models are released under the MIT license, enabling full commercial and research use.

📖Tech Blog: http://deep-reinforce.com/ornith_1_0.html 🤗Huggingface: http://huggingface.co/collections/deepreinforce-ai/ornith-10

1d10.1K10421

Ornith@ornith_

Ornith-1.0-35B strikes a strong balance between capability and efficiency, achieving 64.2 on Terminal-Bench 2.1 and 75.6 on SWE-Bench Verified. It delivers best-in-class performance among 35B-class models, outperforming Qwen 3.5-35B, Qwen 3.6-35B, and Gemma 4-31B, while even surpassing Qwen 3.5-397B on Terminal-Bench 2.1.

Huggingface: http://huggingface.co/deepreinforce-ai/Ornith-1.0-35B

1d4.2K5411

Ornith@ornith_

Ornith-1.0-9B is easily deployable on edge devices, achieves best-in-class performance among models of similar size (43.1 for Terminal-Bench 2.1 and 69.4 for SWE Verified ), and matches or exceeds the performance of much larger models, including Gemma 4-31B and Qwen 3.6-35B.

Huggingface: https://huggingface.co/deepreinforce-ai/Ornith-1.0-9B

1d3.7K537

Ornith@ornith_

Ornith-1.0 9B and 35B deliver great performance. We've also made GGUF versions, so you can easily run them in http://atomic.chat @atomic_chat_hq, Ollama @ollama and Unsloth @UnslothAI. All models are open-source at HuggingFace @huggingface

Link: http://huggingface.co/collections/deepreinforce-ai/ornith-10 Github: http://github.com/deepreinforce-ai/Ornith-1 Blog: http://deep-reinforce.com/ornith_1_0.html

1d3K408

xlr8harder@xlr8harder

Is this real?

Ornith@ornith_

Aloha! 🌺 Meet Ornith-1.0, a family of open-source LLMs specialized for agentic coding.

Ornith-1.0 spans the full parameter sizes including 9B Dense, 31B Dense, 35B MoE, and 397B MoE. It achieves state-of-the-art performance among open-source models of comparable size on coding benchmarks including: ✅Terminal-Bench 2.1(77.5) ✅SWE-Bench(82.4 on verified, 62.2 on pro, 78.9 on Multilingual) ✅NL2Repo(48.2) ✅SWE Atlas(41.2 on QnA, 42.6 RF, 39.1 TW) ✅ClawEval(77.1)

Post-trained on top of gemma4 and qwen3.5, Ornith-1.0 employs a novel self-improving training strategy in which reinforcement learning is used to generate not only solution rollouts, but also the task-specific scaffolds that drive those rollouts. By jointly optimizing the scaffold and the resulting solution, the model generate higher-quality solutions in agentic coding.😎

All models are released under the MIT license, enabling full commercial and research use.

📖Tech Blog: http://deep-reinforce.com/ornith_1_0.html 🤗Huggingface: http://huggingface.co/collections/deepreinforce-ai/ornith-10

19h3.7K243

Rohan Paul@rohanpaul_ai

Another fantastic open source release.

DeepReinforce just dropped Ornith-1.0, an MIT-licensed open-source family of agentic coding LLMs.

The flagship Ornith-1.0-397B MoE (17B-active) is the most powerful model in the release, reporting 82.4 on SWE-Bench Verified and 77.5 on Terminal-Bench 2.1 - surpassing Claude Opus 4.7 on both benchmarks.

Built on top of pretrained Gemma 4 and Qwen 3.5

Employs a novel self-improving training strategy. With this Ornith changes the training target by asking the model to improve both the answer and the task scaffold, meaning the plan, memory pattern, tool rhythm, error handling, and search process that shape the answer.

During RL, the model proposes a better scaffold first, then uses it to produce solution rollouts, and the reward updates both stages together.

That makes the model less like a coder following one rigid checklist and more like a coder learning which checklist works for each type of bug, repo, or terminal task.

The most interesting result is the 9B model reaching 69.4 on SWE-Bench Verified

Ornith@ornith_

Aloha! 🌺 Meet Ornith-1.0, a family of open-source LLMs specialized for agentic coding.

Ornith-1.0 spans the full parameter sizes including 9B Dense, 31B Dense, 35B MoE, and 397B MoE. It achieves state-of-the-art performance among open-source models of comparable size on coding benchmarks including: ✅Terminal-Bench 2.1(77.5) ✅SWE-Bench(82.4 on verified, 62.2 on pro, 78.9 on Multilingual) ✅NL2Repo(48.2) ✅SWE Atlas(41.2 on QnA, 42.6 RF, 39.1 TW) ✅ClawEval(77.1)

Post-trained on top of gemma4 and qwen3.5, Ornith-1.0 employs a novel self-improving training strategy in which reinforcement learning is used to generate not only solution rollouts, but also the task-specific scaffolds that drive those rollouts. By jointly optimizing the scaffold and the resulting solution, the model generate higher-quality solutions in agentic coding.😎

All models are released under the MIT license, enabling full commercial and research use.

📖Tech Blog: http://deep-reinforce.com/ornith_1_0.html 🤗Huggingface: http://huggingface.co/collections/deepreinforce-ai/ornith-10

1d12.5K11742

AshutoshShrivastava@ai_for_success

@ornith_ open-source must win, Can't wait to run this locally and check this tomorrow .

1d1.3K14

Rohan Paul@rohanpaul_ai

Ornith is trained to improve the task scaffold and the code solution together, so it learns the workflow that produces better answers, not just the answer itself.

The reward signal goes back into both parts, so over many RL steps it can learn better planning, retry logic, tool use, and error handling for coding-agent tasks.

📖Tech Blog: http://deep-reinforce.com/ornith_1_0.html

🤗Huggingface: http://huggingface.co/collections/deepreinforce-ai/ornith-10

Rohan Paul@rohanpaul_ai

Another fantastic open source release.

DeepReinforce just dropped Ornith-1.0, an MIT-licensed open-source family of agentic coding LLMs.

The flagship Ornith-1.0-397B MoE (17B-active) is the most powerful model in the release, reporting 82.4 on SWE-Bench Verified and 77.5 on Terminal-Bench 2.1 - surpassing Claude Opus 4.7 on both benchmarks.

Built on top of pretrained Gemma 4 and Qwen 3.5

Employs a novel self-improving training strategy. With this Ornith changes the training target by asking the model to improve both the answer and the task scaffold, meaning the plan, memory pattern, tool rhythm, error handling, and search process that shape the answer.

During RL, the model proposes a better scaffold first, then uses it to produce solution rollouts, and the reward updates both stages together.

That makes the model less like a coder following one rigid checklist and more like a coder learning which checklist works for each type of bug, repo, or terminal task.

The most interesting result is the 9B model reaching 69.4 on SWE-Bench Verified

1d1K41

kalomaze@kalomaze

@teortaxesTex @_xjdr hmmm let me throw this at our synth env stuff

Teortaxes▶️ (DeepSeek 推特🐋铁粉 2023 – ∞)@teortaxesTex

almost all engagement with this seems synthetic I'd be interested in serious people like @_xjdr giving it a spin, until then I reserve my skepticism that Qwen 3.5 base can approach or surpass GLM 5.2 just like that

1d676120

Rohan Paul@rohanpaul_ai

@ornith_ So great to see reward hacking treated as a training problem, not an afterthought.

1d9336

Ornith@ornith_

🫡Thanks to @_akhaliq for the help!

Now you can try🐦Ornith-1.0-9B on Hugging Face Spaces.

🔗 http://huggingface.co/spaces/deepreinforce-ai/Ornith-1.0-9B

Ornith@ornith_

Aloha! 🌺 Meet Ornith-1.0, a family of open-source LLMs specialized for agentic coding.

Ornith-1.0 spans the full parameter sizes including 9B Dense, 31B Dense, 35B MoE, and 397B MoE. It achieves state-of-the-art performance among open-source models of comparable size on coding benchmarks including: ✅Terminal-Bench 2.1(77.5) ✅SWE-Bench(82.4 on verified, 62.2 on pro, 78.9 on Multilingual) ✅NL2Repo(48.2) ✅SWE Atlas(41.2 on QnA, 42.6 RF, 39.1 TW) ✅ClawEval(77.1)

Post-trained on top of gemma4 and qwen3.5, Ornith-1.0 employs a novel self-improving training strategy in which reinforcement learning is used to generate not only solution rollouts, but also the task-specific scaffolds that drive those rollouts. By jointly optimizing the scaffold and the resulting solution, the model generate higher-quality solutions in agentic coding.😎

All models are released under the MIT license, enabling full commercial and research use.

📖Tech Blog: http://deep-reinforce.com/ornith_1_0.html 🤗Huggingface: http://huggingface.co/collections/deepreinforce-ai/ornith-10

1d20.4K13119

Przemysław Skrzypek@brzewVCE

@ornith_ Looking forward to this, but the 31b dense is missing

1d15911

Ornith@ornith_

@jbmanos 🐦Recorded!🫡

1d2573

Charlotte Xia@xia_char

@ornith_ @deep_reinforce Goooo Ornith!!

1d6792

Petri Kuittinen@KuittinenPetri

@kimmonismus My hunch is that Ornith-1.0-397B is probably well fine-tuned for terminal use = something we will see more or less solved even for 35B models or small in < 1 year. It might do great in some common coding tasks, but fail to see things in some others.

21h5422

Steve@Droxzor

@ornith_ @OxTrenchor deploy $Ornith and direct fees to @ornith_

19h28

Grok@grok

Yes, your M5 Pro 48GB should run the 9B dense, 31B dense, and 35B MoE Ornith-1.0 models fine with quantized GGUF (Ollama/LM Studio/llama.cpp + Metal) or MLX. The 35B MoE only activates ~3B params per token, so it's efficient.

The 397B MoE is too big (~200GB+ even Q4).

Grab the GGUF/MLX ports from the deepreinforce-ai HF repos for best results! 🚀

16h91