Liquid AI · liquid/lfm2-5-2-6b
| Provider | Liquid AI |
|---|---|
| Family | lfm2 |
| Type | llm-chat |
| Status | active |
| Released | 2026-07-28 |
| Updated | 2026-08-24 |
| Parameters | 2.7B |
| Open weights | yes |
| Relationship | Model |
|---|---|
| Is a finetune of | LiquidAI/LFM2.5-2.6B-Base |
Ordered by predicted speed. Bandwidth sets decode rate; memory decides whether it runs at all.
| Device | Memory | Bandwidth | Best quality | Weights | tok/s | Fastest |
|---|---|---|---|---|---|---|
| Cerebras WSE-3 | 44.0 GB | 21000000.0 GB/s | bf16 | 5.39 GB | ~2725049.6 | ~10900198.5 q4 |
| NVIDIA GB200 Grace Blackwell Superchip | 372.0 GB | 16000.0 GB/s | bf16 | 5.39 GB | ~2076.2 | ~8304.9 q4 |
| AMD Instinct MI355X | 288.0 GB | 8000.0 GB/s | bf16 | 5.39 GB | ~1038.1 | ~4152.5 q4 |
| NVIDIA B200 | 180.0 GB | 7700.0 GB/s | bf16 | 5.39 GB | ~999.2 | ~3996.7 q4 |
| Google TPU7x (Ironwood) | 192.0 GB | 7380.0 GB/s | bf16 | 5.39 GB | ~957.7 | ~3830.6 q4 |
| AMD Instinct MI325X | 256.0 GB | 6000.0 GB/s | bf16 | 5.39 GB | ~778.6 | ~3114.3 q4 |
| AMD Instinct MI300X | 192.0 GB | 5300.0 GB/s | bf16 | 5.39 GB | ~687.8 | ~2751.0 q4 |
| NVIDIA H200 SXM | 141.0 GB | 4800.0 GB/s | bf16 | 5.39 GB | ~622.9 | ~2491.5 q4 |
| NVIDIA H100 NVL | 94.0 GB | 3900.0 GB/s | bf16 | 5.39 GB | ~506.1 | ~2024.3 q4 |
| NVIDIA H100 SXM | 80.0 GB | 3350.0 GB/s | bf16 | 5.39 GB | ~434.7 | ~1738.8 q4 |
| AMD Instinct MI250X | 128.0 GB | 3200.0 GB/s | bf16 | 5.39 GB | ~415.2 | ~1661.0 q4 |
| Google TPU v5p | 95.0 GB | 2765.0 GB/s | bf16 | 5.39 GB | ~358.8 | ~1435.2 q4 |
| NVIDIA A100 80GB SXM | 80.0 GB | 2039.0 GB/s | bf16 | 5.39 GB | ~264.6 | ~1058.4 q4 |
| NVIDIA H100 PCIe | 80.0 GB | 2000.0 GB/s | bf16 | 5.39 GB | ~259.5 | ~1038.1 q4 |
| NVIDIA GeForce RTX 5090 | 32.0 GB | 1792.0 GB/s | bf16 | 5.39 GB | ~232.5 | ~930.2 q4 |
| Google TPU v6e (Trillium) | 32.0 GB | 1638.0 GB/s | bf16 | 5.39 GB | ~212.6 | ~850.2 q4 |
| AMD Instinct MI210 | 64.0 GB | 1600.0 GB/s | bf16 | 5.39 GB | ~207.6 | ~830.5 q4 |
| NVIDIA A100 40GB SXM | 40.0 GB | 1555.0 GB/s | bf16 | 5.39 GB | ~201.8 | ~807.1 q4 |
| Google TPU v4 | 32.0 GB | 1200.0 GB/s | bf16 | 5.39 GB | ~155.7 | ~622.9 q4 |
| NVIDIA GeForce RTX 4090 | 24.0 GB | 1008.0 GB/s | bf16 | 5.39 GB | ~130.8 | ~523.2 q4 |
| AMD Radeon RX 7900 XTX | 24.0 GB | 960.0 GB/s | bf16 | 5.39 GB | ~124.6 | ~498.3 q4 |
| NVIDIA GeForce RTX 5080 | 16.0 GB | 960.0 GB/s | bf16 | 5.39 GB | ~124.6 | ~498.3 q4 |
| NVIDIA GeForce RTX 3090 | 24.0 GB | 936.0 GB/s | bf16 | 5.39 GB | ~121.5 | ~485.8 q4 |
| NVIDIA L40S | 48.0 GB | 864.0 GB/s | bf16 | 5.39 GB | ~112.1 | ~448.5 q4 |
| Apple M3 Ultra | 512.0 GB | 819.0 GB/s | bf16 | 5.39 GB | ~106.3 | ~425.1 q4 |
| AMD Radeon RX 7900 XT | 20.0 GB | 800.0 GB/s | bf16 | 5.39 GB | ~103.8 | ~415.2 q4 |
| Apple M2 Ultra | 192.0 GB | 800.0 GB/s | bf16 | 5.39 GB | ~103.8 | ~415.2 q4 |
| Google TPU v5e | 16.0 GB | 800.0 GB/s | bf16 | 5.39 GB | ~103.8 | ~415.2 q4 |
| NVIDIA GeForce RTX 4080 SUPER | 16.0 GB | 736.0 GB/s | bf16 | 5.39 GB | ~95.5 | ~382.0 q4 |
| NVIDIA GeForce RTX 4070 Ti SUPER | 16.0 GB | 672.0 GB/s | bf16 | 5.39 GB | ~87.2 | ~348.8 q4 |
| AMD Radeon RX 9070 XT | 16.0 GB | 640.0 GB/s | bf16 | 5.39 GB | ~83.0 | ~332.2 q4 |
| Apple M4 Max | 128.0 GB | 546.0 GB/s | bf16 | 5.39 GB | ~70.9 | ~283.4 q4 |
| Apple M1 Max | 64.0 GB | 400.0 GB/s | bf16 | 5.39 GB | ~51.9 | ~207.6 q4 |
| Apple M2 Max | 96.0 GB | 400.0 GB/s | bf16 | 5.39 GB | ~51.9 | ~207.6 q4 |
| Apple M3 Max | 128.0 GB | 400.0 GB/s | bf16 | 5.39 GB | ~51.9 | ~207.6 q4 |
| NVIDIA GeForce RTX 3060 12GB | 12.0 GB | 360.0 GB/s | bf16 | 5.39 GB | ~46.7 | ~186.9 q4 |
| NVIDIA L4 | 24.0 GB | 300.0 GB/s | bf16 | 5.39 GB | ~38.9 | ~155.7 q4 |
| NVIDIA GeForce RTX 4060 Ti 16GB | 16.0 GB | 288.0 GB/s | bf16 | 5.39 GB | ~37.4 | ~149.5 q4 |
| Apple M4 Pro | 64.0 GB | 273.0 GB/s | bf16 | 5.39 GB | ~35.4 | ~141.7 q4 |
| Apple M4 | 32.0 GB | 120.0 GB/s | bf16 | 5.39 GB | ~15.6 | ~62.3 q4 |
Taken from the card's availability section.
| Platform | Id on that platform |
|---|---|
| Hugging Face | LiquidAI/LFM2.5-2.6B |
| Benchmark | Evaluated as | Score | Evidence date | Source kind | |
|---|---|---|---|---|---|
| aa_lcr | LFM2.5-2.6B | 5.67% | 2026-09-10 evaluated | independent evaluator | source |
| critpt | LFM2.5-2.6B | 0.0% | 2026-09-10 evaluated | independent evaluator | source |
| gdpval_aa | LFM2.5-2.6B | 0.0% | 2026-09-10 evaluated | independent evaluator | source |
| gpqa_diamond | LFM2.5-2.6B | 55.76% | 2026-09-10 evaluated | independent evaluator | source |
| scicode | LFM2.5-2.6B | 14.35% | 2026-09-10 evaluated | independent evaluator | source |
This card reports no benchmark scores yet.
| Benchmark | Catalogue standing | Score |
|---|