4-bit Open on Hugging Face
Parameters
36B
Weights (est.)
≈ 20.4 GB
Mac memory (est.)
32 GB+
Size is estimated from the parameter count and quantisation. Memory allows a fifth extra for context plus 2 GB for the system; long contexts need more.
Run it
With mlx-lm on an Apple silicon Mac:
pip install mlx-lmmlx_lm.generate --model mlx-community/TeleChat3-36B-Thinking-4bit --prompt "Hello"
Details
- Base model
- Tele-AI/TeleChat3-36B-Thinking (quantized)
- Vendor
- Tele-AI
- Task
- Text generation
- Quantisation
- 4-bit
- Licence
- —
- Downloads
- 29 in the last 30 days · 183 all time · 0 likes
- Published
- 2026-01-19 · updated 2026-01-19 (9mo ago)
- Tags
- telechat3, custom_code
Other MLX builds of Tele-AI/TeleChat3-36B-Thinking
| Build | Quantisation | Weights | Mac memory | Downloads |
|---|---|---|---|---|
| TeleChat3-36B-Thinking-4bit (this one) | 4-bit | ≈ 20.4 GB | 32 GB+ | 29 |
| TeleChat3-36B-Thinking-6bit | 6-bit | ≈ 29.4 GB | 48 GB+ | 26 |
| TeleChat3-36B-Thinking-8bit | 8-bit | ≈ 38.5 GB | 48 GB+ | 31 |
| TeleChat3-36B-Thinking-bf16 | bf16 | ≈ 72.5 GB | 96 GB+ | 26 |