Model benchmarks
peculiar-ragdoll/Tiel-Coder-35B-A3B-GGUF-MTP local LLM performance
As of October 2026, peculiar-ragdoll/Tiel-Coder-35B-A3B-GGUF-MTP runs at up to 101.4 tok/s for local inference (best of 5 community benchmark runs across 2 GPUs).
Model size
35B
Peak speed
101.4 tok/s
Average speed
41.7 tok/s
Avg PP
2357.8 tok/s
Min memory
n/a
Max context
262,144 tokens
Avg output / run
8,585 tokens
Avg runtime / run
3m 19s
Avg quality
79.4
Benchmark runs
5
GPUs tested
2
Quality by task
Average LLM-judged quality (0–100) with the run-to-run spread shown as a P5–P95 band, overall and for each benchmark task, across all 5 runs. The low and high columns show how much the judge’s score varies between runs, and need at least two runs to display.
| Task | P5 (low) | Avg | P95 (high) |
|---|---|---|---|
| Overall | 77.5 | 79.4 | 80.8 |
| Agent Workflow | 75.0 | 80.2 | 82.3 |
| Code Generation | 75.2 | 78.7 | 83.9 |
| Role Play & Narrative | 74.9 | 81.9 | 85.6 |
| Research & Analysis | 74.7 | 76.7 | 79.6 |
Performance by hardware and tool
Every hardware/tool/quantization combination peculiar-ragdoll/Tiel-Coder-35B-A3B-GGUF-MTP has been benchmarked on, ranked by peak token generation speed. Last updated October 2026.
| Hardware | Tool | Quant | Best tok/s | Avg tok/s | Memory | Context | Quality | Runs |
|---|---|---|---|---|---|---|---|---|
| Apple M5 Max | Unsloth Studio | UD-Q8_K_XL | 101.4 tok/s | 101.4 tok/s | n/a | 65,536 tokens | 81.0 | 1 |
| Advanced Micro Devices, Inc. [AMD/ATI] HawkPoint1 | llama.cpp | — | 26.9 tok/s | 26.8 tok/s | n/a | 262,144 tokens | 78.9 | 4 |
Benchmark runs
All 5 peculiar-ragdoll/Tiel-Coder-35B-A3B-GGUF-MTP runs submitted so far — expand one for its hardware, quality breakdown and per-scenario detail.
Frequently asked questions
- Is peculiar-ragdoll/Tiel-Coder-35B-A3B-GGUF-MTP good for coding?
- In our benchmarks, peculiar-ragdoll/Tiel-Coder-35B-A3B-GGUF-MTP scores 78.7/100 for coding. It runs at about 41.7 tok/s, so if you want more speed, local-qwen is faster (~60.2 tok/s) and still scores well for coding (84.8/100).
- Is peculiar-ragdoll/Tiel-Coder-35B-A3B-GGUF-MTP good for agentic (tool-using) tasks?
- In our benchmarks, peculiar-ragdoll/Tiel-Coder-35B-A3B-GGUF-MTP scores 80.2/100 for agentic workflows. It runs at about 41.7 tok/s, so if you want more speed, granite-4.2-8b-Q4_K_L is faster (~71.6 tok/s) and still scores well for agentic workflows (92.9/100).
- How fast is peculiar-ragdoll/Tiel-Coder-35B-A3B-GGUF-MTP for local inference?
- Across 5 community benchmark runs, peculiar-ragdoll/Tiel-Coder-35B-A3B-GGUF-MTP reaches up to 101.4 tok/s and averages 41.7 tok/s, with the fastest results on Apple M5 Max.
- Which tools have been used to run peculiar-ragdoll/Tiel-Coder-35B-A3B-GGUF-MTP?
- Benchmarks were submitted using Unsloth Studio, llama.cpp. Results are community-contributed and updated as new runs arrive.