Once you have tasted flight, you will walk the earth with your eyes turned skywards.
SF Bay Area
Joined June 2026
- Tweets580
- Following371
- Followers20.1K
- Likes3K
Pinned Tweet
Aloha!🌺 Meet Ornith-1.5-122B: No Tradeoff Between Strong and Fast
Stronger. Faster. Leaner. Ornith-1.5-122B delivers frontier-level capabilities with exceptional inference efficiency. With superior instruction following, stronger task decomposition, and deeper understanding of real-world workflows, it achieves state-of-the-art performance among open-source models of comparable size, while running on a single NVIDIA DGX card!!
Open weights under the MIT license.
Coding
✅SWE-bench Verified: 0.845
✅TB2.1: 0.813
✅CyberGym: 0.800
✅NL2Repo: 0.512
Agentic
✅HLE: 0.446
✅BrowseComp: 0.785
✅MCP-Atlas: 0.746
✅Toolathlon: 0.708
🤗Hugging Face: huggingface.co/ornith-ai/Orn…
We can’t wait to see what you build with Ornith-1.5-122B.
Ornith retweeted
Run Ornith-1.5-122B on your DGX Spark ✨
Uses my Ornith-1.5-122B-EXL3-EAQ 4 bpw quant. Recipe can easily be adapted for the RTX PRO 6000 too.
- Runs on TensorFold
- 256K context, total 1M KV cache
- 4 concurrent streams by default
- 86 GB total memory
- Uses DFlash2 drafts
Decode speed (thinking off):
- 52 tok/s on prose, single stream
- 103 tok/s on code, single stream
- 90 tok/s on prose, 4 concurrent streams
Prefill is ~2000 tok/s at most context sizes.
Overall I like this model. It feels different from Qwen3.8-Flash-Next and does things its own way. Give it a try!
Get it here:
github.com/MiaAI-Lab/Ornith-…
🐦 Ornith-1.5, including the recent 122B one, are built upon Ornith-1.0, which was developed on top of Qwen3.5 @Alibaba_Qwen with additional continued pretraining, mid-training, and post-training.
🫡 Thanks to the open-weight community!
👀 Please feel free to build your own projects or model variants on top of Ornith.
🐦Ornith-1.5 is built upon Ornith-1.0, which was developed on top of Qwen3.5 @Alibaba_Qwen with additional continued pretraining, mid-training, and post-training.
🫡Thanks to the open-weight community!
👀We are happy and welcome more projects/model variants built on Ornith.
🐦 Now you can serve Ornith-1.5-122B with SGLang
🫡 Shoutout to @sgl_project for their contributions to the open-weight community.
Deployment command👇
Day-0 SGLang support for Ornith-1.5-122B! 🚀
Built for coding and agentic tasks, now ready to serve with SGLang, including reasoning, tool calling, and 262K context configurations.
Thanks to the Ornith team for the collaboration and cookbook contribution!
Get started 👇
docs.sglang.io/cookbook/auto…
Ornith retweeted
Day-0 SGLang support for Ornith-1.5-122B! 🚀
Built for coding and agentic tasks, now ready to serve with SGLang, including reasoning, tool calling, and 262K context configurations.
Thanks to the Ornith team for the collaboration and cookbook contribution!
Get started 👇
docs.sglang.io/cookbook/auto…
Aloha!🌺 Meet Ornith-1.5-122B: No Tradeoff Between Strong and Fast
Stronger. Faster. Leaner. Ornith-1.5-122B delivers frontier-level capabilities with exceptional inference efficiency. With superior instruction following, stronger task decomposition, and deeper understanding of real-world workflows, it achieves state-of-the-art performance among open-source models of comparable size, while running on a single NVIDIA DGX card!!
Open weights under the MIT license.
Coding
✅SWE-bench Verified: 0.845
✅TB2.1: 0.813
✅CyberGym: 0.800
✅NL2Repo: 0.512
Agentic
✅HLE: 0.446
✅BrowseComp: 0.785
✅MCP-Atlas: 0.746
✅Toolathlon: 0.708
🤗Hugging Face: huggingface.co/ornith-ai/Orn…
We can’t wait to see what you build with Ornith-1.5-122B.
🧡Ornith-1.5-122B is now live on Hermes Agent!!!
🫡Huge thanks to @NousResearch for their extraordinary work.
Feel free to set goals for agents, give Ornith-1.5-122B a try, and let us know what you think!
🔗 portal.nousresearch.com/mode…
Ornith-1.5-122B is now available for FREE in Nous Portal for 48 hours
thank you @ornith_
If you are not yet, you will soon be tired of winning
(but your Hermes Agent will never tire)
Ornith retweeted
We made some Unsloth UD3 dynamic quants for Ornith's new 122B release! UD-Q2_K_XL is 44.5GB and retains around 84% top-1% for wikitext general and 89% ish for chat whilst being 82% smaller (BF16 = 249GB)
🥳 Ornith-1.5-122B Unsloth GGUFs are now available!
🧡 Thanks for the awesome work by @UnslothAI
🔗 huggingface.co/unsloth/Ornit…
Ornith retweeted
Ornith-1.5-122B is now available for FREE in Nous Portal for 48 hours
thank you @ornith_
If you are not yet, you will soon be tired of winning
(but your Hermes Agent will never tire)
Aloha!🌺 Meet Ornith-1.5-122B: No Tradeoff Between Strong and Fast
Stronger. Faster. Leaner. Ornith-1.5-122B delivers frontier-level capabilities with exceptional inference efficiency. With superior instruction following, stronger task decomposition, and deeper understanding of real-world workflows, it achieves state-of-the-art performance among open-source models of comparable size, while running on a single NVIDIA DGX card!!
Open weights under the MIT license.
Coding
✅SWE-bench Verified: 0.845
✅TB2.1: 0.813
✅CyberGym: 0.800
✅NL2Repo: 0.512
Agentic
✅HLE: 0.446
✅BrowseComp: 0.785
✅MCP-Atlas: 0.746
✅Toolathlon: 0.708
🤗Hugging Face: huggingface.co/ornith-ai/Orn…
We can’t wait to see what you build with Ornith-1.5-122B.
👀 We tried to optimize coding performance for Ornith-1.5-122B.
Now the model is free on CommandCode for a limited time.
🫡 Please give it a spin and tell us what you think!
commandcode.ai
Ornith retweeted
Ornith-1.5-122B is now live & free in Command Code.
125B params, 262K context, ZDR by default
Open-weight strong and fast agentic coding model
Available for all plans, all subscribers.
🐐
Ornith retweeted
Congrats Ornith team on another amazing open model. 🌺 We made Ornith-1.5-122B GGUFs to enable you to run locally on 64GB RAM setups!
huggingface.co/unsloth/Ornit…
Ornith retweeted
more open models drops!
and the best part, Hermes Agent is in the launch list:
`ollama launch hermes --model ornith-1.5:122b`
Aloha!🌺 Meet Ornith-1.5-122B: No Tradeoff Between Strong and Fast
Stronger. Faster. Leaner. Ornith-1.5-122B delivers frontier-level capabilities with exceptional inference efficiency. With superior instruction following, stronger task decomposition, and deeper understanding of real-world workflows, it achieves state-of-the-art performance among open-source models of comparable size, while running on a single NVIDIA DGX card!!
Open weights under the MIT license.
Coding
✅SWE-bench Verified: 0.845
✅TB2.1: 0.813
✅CyberGym: 0.800
✅NL2Repo: 0.512
Agentic
✅HLE: 0.446
✅BrowseComp: 0.785
✅MCP-Atlas: 0.746
✅Toolathlon: 0.708
🤗Hugging Face: huggingface.co/ornith-ai/Orn…
We can’t wait to see what you build with Ornith-1.5-122B.
🥳 Ornith-1.5-122B Unsloth GGUFs are now available!
🧡 Thanks for the awesome work by @UnslothAI
🔗 huggingface.co/unsloth/Ornit…
Aloha!🌺 Meet Ornith-1.5-122B: No Tradeoff Between Strong and Fast
Stronger. Faster. Leaner. Ornith-1.5-122B delivers frontier-level capabilities with exceptional inference efficiency. With superior instruction following, stronger task decomposition, and deeper understanding of real-world workflows, it achieves state-of-the-art performance among open-source models of comparable size, while running on a single NVIDIA DGX card!!
Open weights under the MIT license.
Coding
✅SWE-bench Verified: 0.845
✅TB2.1: 0.813
✅CyberGym: 0.800
✅NL2Repo: 0.512
Agentic
✅HLE: 0.446
✅BrowseComp: 0.785
✅MCP-Atlas: 0.746
✅Toolathlon: 0.708
🤗Hugging Face: huggingface.co/ornith-ai/Orn…
We can’t wait to see what you build with Ornith-1.5-122B.
Ornith retweeted
I've asked Ornith-1.5-122B to create 100 HTML files
Rules were simple:
- look stunning
- zero repetitive designs
- go full creative mode
I'm gonna be short - this models punches WAY above its weight! Seriously.
Even some latest "frontier" models didn't wow me this much. This is no Opus 5.5, but for a model of THIS size, I'm impressed.
I think there are almost no broken files? Maybe some, but very few.
What surprised me was its creativity - it produced some unique files that other models didn't.
There were SOME repetitive files, not nothing major.
This could be a great model to run on ONE DGX SPARK or RTX 6000 PRO, and you'll have MORE room for other stuff alongside it!
Providing results AS-IS, with no edits!
See the results for yourself (incl. prompts):
miaai-lab.github.io/Ornith-1…
All other 100-HTML file experiments:
mia-ai.net/experiments/100-h…
Some that I liked, and there are quite a lot!
POP ROOM!
miaai-lab.github.io/Ornith-1…
NEXUS - Interactive Galaxy Field
miaai-lab.github.io/Ornith-1…
Flow Field - Ink on Cream
miaai-lab.github.io/Ornith-1…
Ink Bamboo on Rice Paper
miaai-lab.github.io/Ornith-1…
Aurora - Intensity Field
miaai-lab.github.io/Ornith-1…
Atomic Age
miaai-lab.github.io/Ornith-1…
CRYSTALLINE - A Gem Growing From a Single Seed
miaai-lab.github.io/Ornith-1…
Warp - Hyperspace Transit
miaai-lab.github.io/Ornith-1…
QUANTUM · Probability Field
miaai-lab.github.io/Ornith-1…
A WINTER IN GLASS
miaai-lab.github.io/Ornith-1…
METEORA - A Shower of Falling Stars
miaai-lab.github.io/Ornith-1…
Zentangle
miaai-lab.github.io/Ornith-1…
The Dreaming Jellyfish
miaai-lab.github.io/Ornith-1…
Kepler's Playground
miaai-lab.github.io/Ornith-1…
👻 It's time to play with Ornith-1.5-122B on HF 🤗
Thanks @_akhaliq and @huggingface for the support!!!
🔗 huggingface.co/spaces/ornith…
Aloha!🌺 Meet Ornith-1.5-122B: No Tradeoff Between Strong and Fast
Stronger. Faster. Leaner. Ornith-1.5-122B delivers frontier-level capabilities with exceptional inference efficiency. With superior instruction following, stronger task decomposition, and deeper understanding of real-world workflows, it achieves state-of-the-art performance among open-source models of comparable size, while running on a single NVIDIA DGX card!!
Open weights under the MIT license.
Coding
✅SWE-bench Verified: 0.845
✅TB2.1: 0.813
✅CyberGym: 0.800
✅NL2Repo: 0.512
Agentic
✅HLE: 0.446
✅BrowseComp: 0.785
✅MCP-Atlas: 0.746
✅Toolathlon: 0.708
🤗Hugging Face: huggingface.co/ornith-ai/Orn…
We can’t wait to see what you build with Ornith-1.5-122B.
🙌 Ornith-1.5-122B is live on the RunInfra API!!!
Now you can plug the model into agentic workflows via API.
🧡 Thanks @runinfrai for the support!!
🔗 runinfra.ai/inference-api/or…
Ornith 1.5 122B is live on runinfra. launch partner with @ornith_ , live from hour 0
302 tok/s output. 72ms to first reasoning token
$0.20 / 1M input. $0.02 cached. $0.80 output
99.1% cache hit rate on agentic workloads
one of the best models we've served for its size. 10B active and it holds up on real agent work
OpenAI-compatible chat completions and Anthropic-compatible /v1/messages -- drop it into claude code as is
zero data retention by default
runinfra.ai/inference-api/or…
Ornith retweeted
Ornith 1.5 122B is live on runinfra. launch partner with @ornith_ , live from hour 0
302 tok/s output. 72ms to first reasoning token
$0.20 / 1M input. $0.02 cached. $0.80 output
99.1% cache hit rate on agentic workloads
one of the best models we've served for its size. 10B active and it holds up on real agent work
OpenAI-compatible chat completions and Anthropic-compatible /v1/messages -- drop it into claude code as is
zero data retention by default
runinfra.ai/inference-api/or…
🪽Ornith-1.5-122B gets its DFlash2 wings from Day 0!
Huge thanks to @inco_ai for the amazing work on inference speedup.
🔗 huggingface.co/incoai/Ornith…
Congrats to @ornith_ on releasing Ornith-1.5-122B! We trained a DFlash 2 drafter for day-0 support.
It speeds up its generation by up to 3×, and it's lossless!
huggingface.co/incoai/Ornith…
🐦 Shoutout to @atomic_chat_hq for their expertise!
Compressed Ornith-1.5-122B from 132.6GB down to 35.0GB.
🥳 Come and try Ornith 1.5 122B on atomic.chat
Run Ornith 1.5 122B locally on 64GB Mac 🐦🔥
We released our quant ladder for @ornith_'s new 122B model on Hugging Face from 132.6GB down to 35.0GB
For 64GB Macs we recommend the 47.6GB AD-IQ3_XXS-IQ2_S with a top-1% of 86.0% compared to BF16
GGUFs 👇
huggingface.co/AtomicChat/Or…
Ornith retweeted
Congrats to @ornith_ on releasing Ornith-1.5-122B! We trained a DFlash 2 drafter for day-0 support.
It speeds up its generation by up to 3×, and it's lossless!
huggingface.co/incoai/Ornith…
Aloha!🌺 Meet Ornith-1.5-122B: No Tradeoff Between Strong and Fast
Stronger. Faster. Leaner. Ornith-1.5-122B delivers frontier-level capabilities with exceptional inference efficiency. With superior instruction following, stronger task decomposition, and deeper understanding of real-world workflows, it achieves state-of-the-art performance among open-source models of comparable size, while running on a single NVIDIA DGX card!!
Open weights under the MIT license.
Coding
✅SWE-bench Verified: 0.845
✅TB2.1: 0.813
✅CyberGym: 0.800
✅NL2Repo: 0.512
Agentic
✅HLE: 0.446
✅BrowseComp: 0.785
✅MCP-Atlas: 0.746
✅Toolathlon: 0.708
🤗Hugging Face: huggingface.co/ornith-ai/Orn…
We can’t wait to see what you build with Ornith-1.5-122B.