Compare / Nemotron Nano 9B v2 vs LFM2.5 8B-A1B
Nemotron Nano 9B v2 vs LFM2.5 8B-A1B
Updated November 2026
Verdict
Nemotron Nano 9B v2 scores higher on the Small LM Index: 52.9 vs 40.1. On the 6 benchmarks both labs report, Nemotron Nano 9B v2 wins 4 and LFM2.5 8B-A1B wins 2. Both need about 6.2 GB of memory at 4-bit quantization. Neither model has a hosted API on OpenRouter, so you would run both yourself.
Pick Nemotron Nano 9B v2 if…
- you want the strongest model overall
- you need math and reasoning
- you build agents that call tools
- you want the more widely used model (more quantized builds and community support)
Pick LFM2.5 8B-A1B if…
- you need precise instruction following
Specifications
| Nemotron Nano 9B v2 | LFM2.5 8B-A1B | |
|---|---|---|
| Small LM Index | 52.9 (likely 50–56) | 40.1 (likely 37–43) |
| Lab | NVIDIA · United States | Liquid AI · United States |
| Released | 2025-08-18 | 2026-05-28 |
| Parameters | 8.9B (dense) | 8.5B (1B active, MoE) |
| Memory at 4-bit | ~6.2 GB | ~5.9 GB |
| Context window | 128K tokens | 128K tokens |
| Input | text | text |
| Reasoning | Hybrid (thinking on/off) | No thinking |
| License | NVIDIA Open Model | LFM Open License |
| API price / 1M tokens | Self-host only | Self-host only |
| Hugging Face downloads (30 days) | 133.7K | 31.3K |
Benchmarks
Higher is better. Bold marks the winner where both labs report a score. ≈ marks the index model's estimate for a score that wasn't reported, ✓ an independent measurement and * a score measured by another lab.
| Benchmark | Nemotron Nano 9B v2 | LFM2.5 8B-A1B |
|---|---|---|
| MMLU-Pro Knowledge | 74.2✓ | ≈ 63 |
| GPQA Diamond Knowledge | 57.0✓ | 51.3✓ |
| Humanity's Last Exam Reasoning | 4.9✓ | 6.9✓ |
| AIME Math | 69.7✓ | 42.5 |
| LiveCodeBench Coding | 72.4✓ | ≈ 35 |
| τ²-Bench Agentic | 21.9✓ | 16.1✓ |
| IFBench Instruction following | 27.6✓ | 55.6✓ |
| IFEval Instruction following | — | 91.8 |
| BFCL Agentic | — | 48.5 |
| MATH-500 Math | — | 88.8 |
| AA-LCR Long context | 22.7✓ | 0.0✓ |
Scores are mostly reported by each lab with its own settings, so small gaps are noise. See the methodology. Open in the interactive comparison to add more models.
Questions
Which is better, Nemotron Nano 9B v2 or LFM2.5 8B-A1B?
Nemotron Nano 9B v2 scores higher on the Small LM Index: 52.9 vs 40.1. On the 6 benchmarks both labs report, Nemotron Nano 9B v2 wins 4 and LFM2.5 8B-A1B wins 2.
Which is cheaper to run?
Neither model has a hosted API on OpenRouter, so you would run both yourself. Both need about 6.2 GB of memory at 4-bit quantization.
Which has the longer context window?
Both have a 128K-token context window.
Can I use Nemotron Nano 9B v2 and LFM2.5 8B-A1B commercially?
Nemotron Nano 9B v2 is released under NVIDIA Open Model and LFM2.5 8B-A1B under LFM Open License. Apache 2.0 and MIT allow commercial use; other licenses can add conditions, so read the license before shipping a product.