Model benchmarks
dev-model:coding local LLM performance
As of November 2026, dev-model:coding runs at up to 77.9 tok/s for local inference (best of 2 community benchmark runs across 1 GPU).
Model size
Unknown
Peak speed
77.9 tok/s
Average speed
74.5 tok/s
Avg PP
1135.6 tok/s
Min memory
19.2 GB
Max context
131,072 tokens
Avg output / run
34,341 tokens
Avg runtime / run
7m 39s
Avg quality
86.3
Benchmark runs
2
GPUs tested
1
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 2 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 | 84.7 | 86.3 | 87.8 |
| Agent Workflow | 89.4 | 89.4 | 89.4 |
| Code Generation | 79.9 | 83.1 | 86.3 |
| Role Play & Narrative | 81.0 | 85.2 | 89.4 |
| Research & Analysis | 86.1 | 87.3 | 88.5 |
Performance by hardware and tool
Every hardware/tool/quantization combination dev-model:coding has been benchmarked on, ranked by peak token generation speed. Last updated November 2026.
| Hardware | Tool | Quant | Best tok/s | Avg tok/s | Memory | Context | Quality | Runs |
|---|---|---|---|---|---|---|---|---|
| NVIDIA RTX A2000 12GB | OpenAI-compatible | — | 77.9 tok/s | 74.5 tok/s | 19.2 GB | 131,072 tokens | 86.3 | 2 |
Benchmark runs
All 2 dev-model:coding runs submitted so far — expand one for its hardware, quality breakdown and per-scenario detail.
Frequently asked questions
- Is dev-model:coding good for coding?
- In our benchmarks, dev-model:coding scores 83.1/100 for coding. It runs at about 74.5 tok/s, so if you want more speed, ornith-1.5-35b-a3b-apex-mtp-i-quality is faster (~105.2 tok/s) and still scores well for coding (83.6/100).
- Is dev-model:coding good for agentic (tool-using) tasks?
- In our benchmarks, dev-model:coding scores 89.4/100 for agentic workflows. It runs at about 74.5 tok/s, so if you want more speed, Nex-N2.5-mini-APEX-Mini is faster (~88.2 tok/s) and still scores well for agentic workflows (91.3/100).
- How fast is dev-model:coding for local inference?
- Across 2 community benchmark runs, dev-model:coding reaches up to 77.9 tok/s and averages 74.5 tok/s, with the fastest results on NVIDIA RTX A2000 12GB.
- How much memory does dev-model:coding need?
- The leanest observed configuration used about 19.2 GB of memory.
- Which tools have been used to run dev-model:coding?
- Benchmarks were submitted using OpenAI-compatible. Results are community-contributed and updated as new runs arrive.