Model benchmarks
MiniCPM-V-4.7-35B-A3B.APEX-I-NanoPlus-V2.1 local LLM performance
As of October 2026, MiniCPM-V-4.7-35B-A3B.APEX-I-NanoPlus-V2.1 runs at up to 100.4 tok/s for local inference (best of 3 community benchmark runs across 1 GPU).
Model size
35B
Peak speed
100.4 tok/s
Average speed
92.7 tok/s
Avg PP
1136.0 tok/s
Min memory
16.9 GB
Max context
131,072 tokens
Avg output / run
9,149 tokens
Avg runtime / run
1m 38s
Avg quality
76.1
Benchmark runs
3
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 3 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 | 75.0 | 76.1 | 77.0 |
| Agent Workflow | 81.8 | 84.6 | 87.1 |
| Code Generation | 53.9 | 61.8 | 66.9 |
| Role Play & Narrative | 73.2 | 79.9 | 87.1 |
| Research & Analysis | 77.4 | 78.1 | 78.6 |
Performance by hardware and tool
Every hardware/tool/quantization combination MiniCPM-V-4.7-35B-A3B.APEX-I-NanoPlus-V2.1 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 |
|---|---|---|---|---|---|---|---|---|
| NVIDIA GeForce RTX 5060 | llama.cpp | — | 100.4 tok/s | 92.7 tok/s | 16.9 GB | 131,072 tokens | 76.1 | 3 |
Benchmark runs
All 3 MiniCPM-V-4.7-35B-A3B.APEX-I-NanoPlus-V2.1 runs submitted so far — expand one for its hardware, quality breakdown and per-scenario detail.
Frequently asked questions
- Is MiniCPM-V-4.7-35B-A3B.APEX-I-NanoPlus-V2.1 good for coding?
- In our benchmarks, MiniCPM-V-4.7-35B-A3B.APEX-I-NanoPlus-V2.1 scores 61.8/100 for coding. It runs at about 92.7 tok/s, so if you want more speed, Ornith-1.5-35B-A3B.APEX-I-NanoPlus is faster (~110.5 tok/s) and still scores well for coding (81.1/100).
- Is MiniCPM-V-4.7-35B-A3B.APEX-I-NanoPlus-V2.1 good for agentic (tool-using) tasks?
- In our benchmarks, MiniCPM-V-4.7-35B-A3B.APEX-I-NanoPlus-V2.1 scores 84.6/100 for agentic workflows. It runs at about 92.7 tok/s, so if you want more speed, cyber-tiel-coder-35b-a3b-mtp is faster (~115.9 tok/s) and still scores well for agentic workflows (86.4/100).
- How fast is MiniCPM-V-4.7-35B-A3B.APEX-I-NanoPlus-V2.1 for local inference?
- Across 3 community benchmark runs, MiniCPM-V-4.7-35B-A3B.APEX-I-NanoPlus-V2.1 reaches up to 100.4 tok/s and averages 92.7 tok/s, with the fastest results on NVIDIA GeForce RTX 5060.
- How much memory does MiniCPM-V-4.7-35B-A3B.APEX-I-NanoPlus-V2.1 need?
- The leanest observed configuration used about 16.9 GB of memory.
- Which tools have been used to run MiniCPM-V-4.7-35B-A3B.APEX-I-NanoPlus-V2.1?
- Benchmarks were submitted using llama.cpp. Results are community-contributed and updated as new runs arrive.