Canopy Labs
company · United StatesScores
1 product on the map — 1 open-ish.
Openness
3 medium confidence- weights
- open (finetuned and pretrained 3B checkpoints on the Hub, behind an automatically approved form)
- data
- described (more than 100k hours of English speech for the pretrained model; not released)
- code
- open (pretraining and fine-tuning scripts with configs in the repository)
- license
- Apache-2.0 (declared on both model cards)+Llama-3.2-Community-License (both cards name Llama 3.2 3B Instruct as the base model, and its license travels with derivative weights)
- research-release
- multilingual-family (2025 research-preview pairs for other languages; the pretrained models are tagged llama3.2 and the finetuned ones apache-2.0)
The weights download after an automatic form and are labeled Apache-2.0, but they are built on Meta's Llama 3.2, whose community license, with its cap on the very largest services, carries over to the model. The training and fine-tuning scripts are published, while the English speech it was trained on is only described.
- https://huggingface.co/api/models?author=canopylabs&limit=100 recorded 2026-11-12
Hub listing: the *-pretrain-research_release models are tagged "license:llama3.2" and the *-ft-research_release models "license:apache-2.0".
- https://huggingface.co/api/models/canopylabs/orpheus-3b-0.1-ft recorded 2026-11-12
Hub record: "gated":"auto", cardData "license":"apache-2.0", "base_model": ["meta-llama/Llama-3.2-3B-Instruct", "canopylabs/orpheus-3b-0.1-pretrained"], with model-00001-of-00004.safetensors in the file list.
- https://huggingface.co/api/models/canopylabs/orpheus-3b-0.1-pretrained recorded 2026-11-12
Hub record: "gated":"auto", cardData "license":"apache-2.0", "base_model": ["meta-llama/Llama-3.2-3B-Instruct"].
- https://raw.githubusercontent.com/canopyai/Orpheus-TTS/HEAD/LICENSE recorded 2026-11-12
Code LICENSE body: "Apache License" / "Version 2.0, January 2004".
- https://raw.githubusercontent.com/canopyai/Orpheus-TTS/HEAD/README.md recorded 2026-11-12
README: "Orpheus TTS is a SOTA open-source text-to-speech system built on the Llama-3b backbone"; "Pretrained - Our base model trained on 100k+ hours of English speech data"; "Finetune Model" runs "the training script in finetune/train.py"; a "Pretrain Model" section; "[4/2025] We release a family of multilingual models ... in a research preview".
- https://raw.githubusercontent.com/meta-llama/llama-models/main/models/llama3_2/LICENSE recorded 2026-11-12
"LLAMA 3.2 COMMUNITY LICENSE AGREEMENT": "If you distribute or make available the Llama Materials (or any derivative works thereof) ... you shall (A) provide a copy of this Agreement"; section 2 requires a license from Meta above "700 million monthly active users".
Adoption
2 high confidenceHugging Face downloads of the finetuned and pretrained English checkpoints over the trailing month; the multilingual research models are not counted.
- https://huggingface.co/api/models?author=canopylabs&limit=100 recorded 2026-11-12
"downloads":44920 for canopylabs/orpheus-3b-0.1-ft and 752 for canopylabs/orpheus-3b-0.1-pretrained.
Capability
1 medium confidenceOrpheus has no listener rating and publishes no results on a public test set, so nothing places it beside the rated synthesizers. It sits with MeloTTS, another popular open voice with no published evaluation.
- https://artificialanalysis.ai/text-to-speech/leaderboard recorded 2026-11-12
The 104-row leaderboard has no Orpheus entry.
- https://raw.githubusercontent.com/canopyai/Orpheus-TTS/HEAD/README.md recorded 2026-11-12
README "Abilities": "**Zero-Shot Voice Cloning**"; "**Low Latency**: ~200ms streaming latency for realtime applications"; no evaluation section.