Can I run it?

Can I run Cyber Tiel Coder 35B A3B GGUF MTP on Apple M4 Max (40-core GPU)?

Runs

Runs fully in unified memory on your Apple M4 Max (40-core GPU) at Q6_K, 8K context, about 114–156 tok/s. Weights 27.1 GB plus 0.19 GB of KV cache — 27.9 GB of 36.0 GB usable. Up to 256K context still fits.

Best quantQ6_K est.

Working context8K

Max that fits256K

Speed114–156 tok/s

Weights27.1 GB

KV cache0.19 GB

Total27.9 GB

Usable36.0 GB

Cyber Tiel Coder 35B A3B GGUF MTP details · Apple M4 Max (40-core GPU) page · How this estimate is made

Run commands

llama.cpp

# Quant size is estimated — the file may not be published yet
llama-server -hf peculiar-ragdoll/Cyber-Tiel-Coder-35B-A3B-GGUF-MTP:Q6_K -c 8192 -ngl 40

No matching name in the Ollama library snapshot.

LM Studio

# Quant size is estimated — the file may not be published yet
lms get peculiar-ragdoll/Cyber-Tiel-Coder-35B-A3B-GGUF-MTP
# Then load quant Q6_K

Assumptions

Estimate, not a benchmark. Decode speed uses a memory-bandwidth model at 55–75% of peak.
KV cache is FP16. Engine overhead assumes llama.cpp (0.3 GB + 1% of the weights). Ollama is closer to 0.9 GB + 3%, LM Studio closer to 1 GB + 3%.
Q6_K size is estimated from parameters × bits-per-weight because a GGUF file size was not available.
Unified memory uses 75% for the GPU by default, and always leaves about 5 GB for the OS.
Speed is quoted at 8K context, not at the maximum context that still fits.

Larger models that also fit

Can I run Qwen3.8 35B A3B Distill on Apple M4 Max (40-core GPU)? — Runs, Q6_K, 114–155 tok/s
Can I run Qwen3 Coder Next on Apple M4 Max (40-core GPU)? — Tight fit, UD-Q3_K_XL, 43–59 tok/s
Can I run Llama 4 Scout 17B 16E Instruct on Apple M4 Max (40-core GPU)? — Tight fit, UD-IQ1_M, 33–45 tok/s

Updated 2026-10-24. Profile on this page is the device default.