Model benchmarks
IsValorum/Nex-N2.5-mini-APEX-I-MiniPlus-V2.1-GGUF local LLM performance
As of October 2026, IsValorum/Nex-N2.5-mini-APEX-I-MiniPlus-V2.1-GGUF runs at up to 135.9 tok/s for local inference (best of 4 community benchmark runs across 3 GPUs).
Model size
35B
Peak speed
135.9 tok/s
Average speed
95.3 tok/s
Avg PP
1086.7 tok/s
Min memory
24.4 GB
Max context
65,536 tokens
Avg output / run
38,414 tokens
Avg runtime / run
7m 15s
Avg quality
77.8
Benchmark runs
4
GPUs tested
3
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 4 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 | 62.2 | 77.8 | 84.8 |
| Agent Workflow | 74.3 | 83.1 | 90.0 |
| Code Generation | 10.6 | 58.6 | 83.3 |
| Role Play & Narrative | 80.4 | 86.1 | 90.4 |
| Research & Analysis | 81.5 | 83.5 | 86.1 |
Performance by hardware and tool
Every hardware/tool/quantization combination IsValorum/Nex-N2.5-mini-APEX-I-MiniPlus-V2.1-GGUF 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 |
|---|---|---|---|---|---|---|---|---|
| AMD Radeon RX 7900 XTX | llama.cpp | — | 135.9 tok/s | 135.7 tok/s | n/a | 65,536 tokens | 84.2 | 2 |
| NVIDIA GeForce RTX 3070 Ti | Unsloth Studio | Nex-N2.5-mini.APEX-I-MiniPlus-V2.1 | 56.7 tok/s | 56.7 tok/s | n/a | 8,192 tokens | 58.5 | 1 |
| NVIDIA GeForce RTX 3060 | llama.cpp | — | 53.3 tok/s | 53.3 tok/s | 24.4 GB | 16,384 tokens | 84.5 | 1 |
Benchmark runs
All 4 IsValorum/Nex-N2.5-mini-APEX-I-MiniPlus-V2.1-GGUF runs submitted so far — expand one for its hardware, quality breakdown and per-scenario detail.
Frequently asked questions
- Is IsValorum/Nex-N2.5-mini-APEX-I-MiniPlus-V2.1-GGUF good for coding?
- In our benchmarks, IsValorum/Nex-N2.5-mini-APEX-I-MiniPlus-V2.1-GGUF scores 58.7/100 for coding. It runs at about 95.3 tok/s, so if you want more speed, Qwen3.6-35B-A3B-Q4_K_XL is faster (~156.9 tok/s) and still scores well for coding (82.2/100).
- Is IsValorum/Nex-N2.5-mini-APEX-I-MiniPlus-V2.1-GGUF good for agentic (tool-using) tasks?
- In our benchmarks, IsValorum/Nex-N2.5-mini-APEX-I-MiniPlus-V2.1-GGUF scores 83.1/100 for agentic workflows. It runs at about 95.3 tok/s, so if you want more speed, swift-1.5-qwen3.8-flash-next-iq3_s is faster (~259.0 tok/s) and still scores well for agentic workflows (88.9/100).
- How fast is IsValorum/Nex-N2.5-mini-APEX-I-MiniPlus-V2.1-GGUF for local inference?
- Across 4 community benchmark runs, IsValorum/Nex-N2.5-mini-APEX-I-MiniPlus-V2.1-GGUF reaches up to 135.9 tok/s and averages 95.3 tok/s, with the fastest results on AMD Radeon RX 7900 XTX.
- How much memory does IsValorum/Nex-N2.5-mini-APEX-I-MiniPlus-V2.1-GGUF need?
- The leanest observed configuration used about 24.4 GB of memory (quantizations tested: Nex-N2.5-mini.APEX-I-MiniPlus-V2.1).
- Which tools have been used to run IsValorum/Nex-N2.5-mini-APEX-I-MiniPlus-V2.1-GGUF?
- Benchmarks were submitted using Unsloth Studio, llama.cpp. Results are community-contributed and updated as new runs arrive.