Compare / MiniCPM-V 4.6 vs LFM2.5 1.2B Instruct
MiniCPM-V 4.6 vs LFM2.5 1.2B Instruct
Updated November 2026
Verdict
MiniCPM-V 4.6 scores higher on the Small LM Index: 27.0 vs 25.4, although their likely ranges overlap, so the lead is not certain. On the 5 benchmarks both labs report, MiniCPM-V 4.6 wins 2 and LFM2.5 1.2B Instruct wins 3. Both need about 1.4 GB of memory at 4-bit quantization. Neither model has a hosted API on OpenRouter, so you would run both yourself.
Pick MiniCPM-V 4.6 if…
- you want the strongest model overall
- you build agents that call tools
- you work with long documents
- you need a permissive license (Apache-2.0)
- you need image input
Pick LFM2.5 1.2B Instruct if…
- you need precise instruction following
Specifications
| MiniCPM-V 4.6 | LFM2.5 1.2B Instruct | |
|---|---|---|
| Small LM Index | 27.0 (likely 19–39) | 25.4 (likely 23–28) |
| Lab | OpenBMB (MiniCPM) · China | Liquid AI · United States |
| Released | 2026-04-13 | 2026-01-06 |
| Parameters | 1.3B (dense) | 1.2B (dense) |
| Memory at 4-bit | ~1.4 GB | ~1.3 GB |
| Context window | 262K tokens | 32K tokens |
| Input | text, image, video | text |
| Reasoning | No thinking | No thinking |
| License | Apache-2.0 | LFM Open License |
| API price / 1M tokens | Self-host only | Self-host only |
| Hugging Face downloads (30 days) | 196.5K | 120K |
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 | MiniCPM-V 4.6 | LFM2.5 1.2B Instruct |
|---|---|---|
| MMLU-Pro Knowledge | ≈ 50 | 44.4 |
| GPQA Diamond Knowledge | 30.5✓ | 32.6✓ |
| Humanity's Last Exam Reasoning | 5.1✓ | 6.7✓ |
| AIME Math | ≈ 17 | 14.0 |
| τ²-Bench Agentic | 87.7✓ | 10.8✓ |
| IFBench Instruction following | 26.7✓ | 43.8✓ |
| IFEval Instruction following | — | 86.2 |
| BFCL Agentic | — | 49.1 |
| AA-LCR Long context | 7.0✓ | 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, MiniCPM-V 4.6 or LFM2.5 1.2B Instruct?
MiniCPM-V 4.6 scores higher on the Small LM Index: 27.0 vs 25.4, although their likely ranges overlap, so the lead is not certain. On the 5 benchmarks both labs report, MiniCPM-V 4.6 wins 2 and LFM2.5 1.2B Instruct wins 3.
Which is cheaper to run?
Neither model has a hosted API on OpenRouter, so you would run both yourself. Both need about 1.4 GB of memory at 4-bit quantization.
Which has the longer context window?
MiniCPM-V 4.6 has the longer context window: 262K tokens against 32K.
Can I use MiniCPM-V 4.6 and LFM2.5 1.2B Instruct commercially?
MiniCPM-V 4.6 is released under Apache-2.0 and LFM2.5 1.2B Instruct 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.
Related comparisons
- LFM3 1.7B vs MiniCPM-V 4.6
- MiniCPM-V 4.6 vs Spark-X2.6 1.7B
- EXAONE 4.0 1.2B vs MiniCPM-V 4.6
- Llama 3.2 1B vs MiniCPM-V 4.6
- LFM2.5 1.2B Thinking vs MiniCPM-V 4.6
- EXAONE 4.0 1.2B vs LFM2.5 1.2B Instruct
- LFM2.5 1.2B Instruct vs LFM2.5 1.2B Thinking
- LFM2.5 1.2B Instruct vs MiniCPM5 1B
- LFM2.5 1.2B Instruct vs Llama 3.2 1B
- LFM2.5 1.2B Instruct vs Qwen3.5 0.8B