peculiar-ragdoll
Tiel Coder 35B A3B
Tiel Coder 35B A3B is a 34.7B mixture-of-experts qwen35moe model from peculiar-ragdoll, with about 3.0B parameters active per token. Configured context length is 262,144 tokens. License: mit. Quant sizes are the GGUF file bytes published on Hugging Face when a file exists.
Parameters34.7B / 3.0B active
Context256K
Attentionhybrid
Licensemit
40 layers · 2 KV heads · head dim 256 · 30 linear-attention layers · GGUF repo · config
Quant file sizes
| Quant | File size | Source |
|---|---|---|
| UD-Q8_K_XL | 35.8 GB | Hugging Face file bytes |
| UD-Q6_K_XL | 29.7 GB | Hugging Face file bytes |
| UD-Q5_K_XL | 24.8 GB | Hugging Face file bytes |
| UD-Q4_K_XL | 20.8 GB | Hugging Face file bytes |
| UD-Q4_K_S | 19.5 GB | Hugging Face file bytes |
| UD-IQ4_XS | 16.5 GB | Hugging Face file bytes |
| UD-Q3_K_XL | 15.7 GB | Hugging Face file bytes |
| UD-IQ3_XXS | 12.3 GB | Hugging Face file bytes |
| UD-Q2_K_XL | 11.4 GB | Hugging Face file bytes |
On common machines
| Device | Verdict | Quant | Speed |
|---|---|---|---|
| GeForce RTX 4060 Ti 8GB | Expert offload | UD-Q4_K_XL | 74–101 |
| GeForce RTX 3060 12GB | Expert offload | UD-Q4_K_XL | 93–126 |
| GeForce RTX 4060 Ti 16GB | Tight fit | UD-IQ3_XXS | 121–165 |
| GeForce RTX 4090 | Tight fit | UD-Q4_K_XL | 264–359 |
| GeForce RTX 5090 | Tight fit | UD-Q5_K_XL | 399–544 |
| Apple M4 Pro | Tight fit | UD-IQ4_XS | 88–120 |
| Apple M5 | Won't fit | UD-Q2_K_XL | — |
| CPU only · DDR5-5600 | Runs on CPU | UD-Q4_K_XL | 23–32 |
Smallest devices that fit it fully
- GeForce RTX 5080 — Tight fit, UD-IQ3_XXS, 403–549 tok/s
- GeForce RTX 5070 Ti — Tight fit, UD-IQ3_XXS, 376–512 tok/s
- GeForce RTX 5060 Ti 16GB — Tight fit, UD-IQ3_XXS, 188–256 tok/s
- GeForce RTX 4080 Super — Tight fit, UD-IQ3_XXS, 309–421 tok/s
- GeForce RTX 4080 — Tight fit, UD-IQ3_XXS, 301–410 tok/s
- GeForce RTX 4070 Ti Super — Tight fit, UD-IQ3_XXS, 282–384 tok/s
Run it
Commands for the GeForce RTX 4060 Ti 16GB estimate (UD-IQ3_XXS).
llama.cpp
llama-server -hf peculiar-ragdoll/Tiel-Coder-35B-A3B-GGUF:UD-IQ3_XXS -c 8192 -ngl 40
No matching name in the Ollama library snapshot.
LM Studio
lms get peculiar-ragdoll/Tiel-Coder-35B-A3B-GGUF # Then load quant UD-IQ3_XXS