LocalClaw/LLM/Ornith 1.0 35B GGUF
Open-weight MoE

Ornith 1.0 35B GGUF

DeepReinforce Ornith 1.0 mid-size GGUF release for agentic coding. The Q4_K_M build is listed around 21.2GB, making it a realistic 32GB+ local model compared with the 397B server-grade version.

32 GB power user 32 GB RAM Q4_K_M Agentic coding on 32GB+ local machines
Run with LocalClaw Compare all models
Parameters
35B MoE
Minimum RAM
32 GB
Model size
21.2 GB
Quantization
Q4_K_M

Can Ornith 1.0 35B GGUF run locally?

Ornith 1.0 35B GGUF belongs on 32 GB machines when you want stronger quality without jumping to server hardware.

Search for Ornith-1.0-35B-GGUF in LM Studio or another GGUF-compatible runtime.

deepreinforce-ai/Ornith-1.0-35B-GGUF
chatcodereasoningqualityagentictool-callinggeneral

Install path

01
Check RAM fitMinimum 32 GB RAM. Start with the Q4_K_M quant.
02
Load the modelSearch Ornith-1.0-35B-GGUF in LM Studio.
03
Control locallyUse LocalClaw to manage models, agents, chat, channels and scheduled OpenClaw work.

Strengths

  • Practical GGUF packaging for local runtimes
  • Q4_K_M build is listed around 21.2GB on Hugging Face
  • Strong fit for 32GB+ machines that want agentic coding behavior
  • Better quality ceiling than the compact 9B variant
  • MIT licensed release

Limitations

  • Too heavy for most 16GB laptops once OS and runtime overhead are included
  • Still early; real-world coding quality should be validated against Qwen, GLM and DeepSeek alternatives
  • Q8 and BF16 variants move into high-memory territory
  • Not as easy to run as mainstream 14B class models

Best use cases

  • Agentic coding on 32GB+ local machines
  • OpenClaw experiments with a stronger coding model
  • Local tool-calling research
  • Comparing new MoE coding models
  • Users who want Ornith without server-grade 397B hardware

Capability profile

speed
4
quality
8
coding
9
reasoning
8

Technical notes

Developer
DeepReinforce
License
MIT
Context window
Unknown tokens
Architecture
Mid-size Ornith 1.0 GGUF release using a qwen35moe architecture tag on Hugging Face. It is the practical middle ground between the small 9B model and the server-grade 397B model.

This model fits these next steps

Hardware fit is based on LocalClaw's RAM tier, model size and quantization metadata. Always leave memory headroom for your OS and runtime.

Comfortable headroomMac mini M4 Pro 48GBMobile workstationMacBook Pro M4 Max 36GBPower-user picks32GB RAM guide

Similar models to compare

Qwen 3 (32B) 32BQwen 3 Coder (30B) 30BDeepSeek R1 Distill (32B) 32BOrnith 1.0 9B GGUF 9B

Where to go next