cd /news/large-language-models/unsloth-shrinks-ornith-1-5-122b-to-f… · home › topics › large-language-models › article
[ARTICLE · art-171369] src=alphasignal.ai ↗ pub= topic=large-language-models verified=true sentiment=↑ positive

Unsloth Shrinks Ornith-1.5-122B to Fit on a Single 48GB GPU

Unsloth has published dynamic GGUF quantizations of Ornith AI's Ornith-1.5-122B, a 122-billion-parameter mixture-of-experts reasoning and coding model, with builds ranging from a 40.4 GB UD-IQ1_M file to a 249 GB BF16 reference, so the model can run on a single 80GB GPU at Q4 or on high-memory workstations and CPU-offload systems via llama.cpp-compatible runtimes. Ornith AI reports the model, built on the Qwen3.5 lineage with a self-improvement RL loop, scores 76.9 on SWE-bench Pro and 81.2 on TerminalBench 2.1, beating much larger models, though the release notes show narrower or negative margins on TerminalBench and Humanity's Last Exam. The model ships MIT licensed with built-in tool calling, a native 262K context extendable to roughly 1M tokens via YaRN RoPE scaling factor 4.0, and support for vLLM, SGLang, llama.cpp and Ollama.

read2 min views4 publishedNov 17, 2026
Unsloth Shrinks Ornith-1.5-122B to Fit on a Single 48GB GPU
Image: Alphasignal (auto-discovered)

Unsloth's dynamic GGUF quants shrink Ornith-1.5-122B down to a 40GB file, letting a frontier-class reasoning and coding MoE fit on a single 80GB GPU.

- Unsloth released dynamic GGUF quants of [Ornith-1.5-122B](https://huggingface.co/unsloth/Ornith-1.5-122B-GGUF?utm_source=alphasignal) , a 122B MoE reasoning and coding model.
- Base model from [Ornith AI](https://huggingface.co/ornith-ai/Ornith-1.5-122B?utm_source=alphasignal) is built on Qwen3.5 lineage with self-improvement RL loop.
  • Scores 76.9 on SWE-bench Pro and 81.2 on TerminalBench 2.1, beating much larger models.
  • Quants range from 40.4 GB (UD-IQ1_M) to 249 GB (BF16); Q4 fits a single 80GB GPU.
  • Native 262K context, extendable to ~1M tokens via YaRN RoPE scaling factor 4.0.
  • MIT licensed with built-in tool calling; works with vLLM, SGLang, llama.cpp, and Ollama.

Unsloth has published GGUF quantizations of Ornith AI’s Ornith-1.5-122B, a 122-billion-parameter mixture-of-experts reasoning model. The GGUF collection ranges from a 40.4 GB low-bit build to a 249 GB BF16 reference, making the model accessible through llama.cpp-compatible runtimes on high-memory workstations or systems using CPU offload.

The release matters primarily for coding and agent workloads. Ornith reports strong results on SWE-bench Pro, TerminalBench, BrowseComp, and WideSearch, while Unsloth’s quantizations provide several tradeoffs among output quality, memory use, context length, and throughput.

Three training loops move together #

Ornith-1.5 follows Ornith-1.0, whose lineage includes Qwen3.5 and Gemma4 plus continued pretraining, mid-training, and post-training. The 122B checkpoint is the smallest announced member of the new family, and Ornith says a mobile-oriented variant is planned.

The new training method jointly optimizes three components that earlier iterations treated more narrowly:

  1. Task generation: producing new problems for the model to solve instead of depending entirely on a fixed task set.
  2. Scaffold construction: selecting the prompts, tools, control logic, and intermediate steps used to attempt each task.
  3. Solution rollouts: improving the model’s actions and answers through reinforcement learning.

This loop lets the training system generate tasks, test different solution strategies, and use the resulting rewards to update the policy. Ornith describes its reward design and evaluation harnesses in the release notes.

Strong coding scores, uneven lead #

Ornith’s release reports a large advantage on SWE-bench Pro and competitive agentic scores, while TerminalBench and Humanity’s Last Exam show narrower or negative margins. The benchmarks cover different capabilities: TerminalBench measures command-line work, SWE-bench Pro uses repository-level software issues, BrowseComp tests difficult information retrieval, WideSearch evaluates broad web research, and HLE targets hard knowledge and reasoning tasks.

This story is for Pro members #

You've reached the end of the free preview. Upgrade to AlphaSignal Pro to read the full article - and everything else behind the paywall.

── more in #large-language-models 4 stories · sorted by recency
── more on @unsloth 3 stories trending now
sponsored brought to you by zahid.host 4,200+ EU-deployed projects
reading about agents? ship yours in a single git push.

Run your AI side-project on zahid.host

EU-based hosting, git-push deploys, automatic HTTPS, no cold starts. Free tier with a custom domain — perfect for shipping the agent you just read about.

$git push zahid main
→ Live at https://your-agent.zahid.host ✓
Get free account → Pricing
from €0/mo · no card required
LIVE [news/unsloth-shrinks-orni…] indexed:0 read:2min 2026-11-17 · —