Ornith 1.0 31B
- Context
- 262Ktokens
- Released
- 3mo ago
- Size
- 36Bparameters
Macaron-V1-Tall is a text-generation model from Mind Lab, released alongside sibling Macaron-V1-Venti in the Macaron-V1 family and built on the Qwen3.6-35B-A3B base. Its release centers on a Mixture-of-LoRA (MoL) design in which an L0 router sends each new user request to one of four LoRA specialists covering chat, personal-agent, coding, and Generative UI work. The model is built for personal intelligence, tool use, repository-level coding, and code-native Generative UI.
- Multimodal model
- Open weights · 36B
- Accepts images
- Reasoning
- Tools
How it compares
Additional source evidence
- Source model
- mindlab/macaron-v1-tall
- Last updated
- 2026-07-21
- Model context
- 262K
- Model output limit
- 33K
- Open weights
- Yes
- File attachments
- Yes
- Reasoning
- Yes
- Tool calling
- Yes
- Temperature control
- Yes
Downloadable weights
Complete source record (JSON)
{
"id": "mindlab/macaron-v1-tall",
"name": "Macaron V1 Tall",
"description": "Qwen3.6-based Mixture-of-LoRA model with specialists for chat, personal agents, coding, and generative UI",
"last_updated": "2026-07-21",
"attachment": true,
"reasoning": true,
"tool_call": true,
"temperature": true,
"modalities": {
"input": [
"text",
"image"
],
"output": [
"text"
]
},
"limit": {
"context": 262144,
"output": 32768
},
"release_date": "2026-07-21",
"open_weights": true,
"license": "MIT",
"weights": [
{
"label": "Hugging Face",
"url": "https://huggingface.co/mindlab-research/Macaron-V1-Tall"
}
]
}Frequently asked questions
Running locally
Macaron V1 Tall is estimated to need 25.6 GB of VRAM at Q4, 44.4 GB at Q8 and 83.2 GB at FP16. All figures are estimates from the 36B parameter count, with weights plus about 15% and 0.5 GB for a short context.
No. Estimates say Macaron V1 Tall does not fit on a 24 GB, 16 GB, 12 GB or 8 GB GPU at Q4, Q8 or FP16.
Yes, but with a large memory config. Q4 is estimated at 25.6 GB and, assuming macOS lets the GPU use about 75% of unified memory, the smallest common Mac memory size that fits is 36 GB.
Privacy and data
Yes. Macaron V1 Tall has open weights, so you can self-host it for full control of your data.
Capabilities and specs
262K tokens (262,144). Maximum output is 33K tokens (32,768).
No, Macaron V1 Tall reads only text and images. It does not take audio, video or PDFs.
Yes, tool calling (function calling) is supported.
Yes, Macaron V1 Tall supports a reasoning or thinking mode.
Quality and benchmarks
Matched on model kind and release date with no comparable price, the closest alternatives are Ornith 1.0 31B and Ornith 1.0 35B by DeepReinforce, plus North Mini Code by Cohere.
Ornith 1.0 31BMacaron V1 Tall vs Ornith 1.0 31BOrnith 1.0 35BMacaron V1 Tall vs Ornith 1.0 35BNorth Mini CodeMacaron V1 Tall vs North Mini Code
Licensing and access
Macaron V1 Tall is open weights. Its license is MIT.
Yes. Macaron V1 Tall is licensed under MIT, a permissive license that allows commercial use.
The weights are available in the Hugging Face repository mindlab-research/Macaron-V1-Tall.
Basics
Macaron V1 Tall was made by Mind Lab and released on July 21, 2026.
Yes, Macaron V1 Tall has open weights, so the model can be downloaded and used freely.
Generated from Vertical Tabs data, updated November 11, 2026.
Discussion
No comments yet
Be the first to share your thoughts.