IBM · ibm/granite-4-2-8b
| Provider | IBM |
|---|---|
| Family | granite |
| Type | llm-chat |
| Status | active |
| Released | 2026-08-07 |
| Updated | 2026-09-04 |
| Parameters | 8.8B |
| Open weights | yes |
| Relationship | Model |
|---|---|
| Is a finetune of | ibm-granite/granite-4.1-8b-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 | 17.58 GB | ~836026.0 | ~3344103.9 q4 |
| NVIDIA GB200 Grace Blackwell Superchip | 372.0 GB | 16000.0 GB/s | bf16 | 17.58 GB | ~637.0 | ~2547.9 q4 |
| AMD Instinct MI355X | 288.0 GB | 8000.0 GB/s | bf16 | 17.58 GB | ~318.5 | ~1273.9 q4 |
| NVIDIA B200 | 180.0 GB | 7700.0 GB/s | bf16 | 17.58 GB | ~306.5 | ~1226.2 q4 |
| Google TPU7x (Ironwood) | 192.0 GB | 7380.0 GB/s | bf16 | 17.58 GB | ~293.8 | ~1175.2 q4 |
| AMD Instinct MI325X | 256.0 GB | 6000.0 GB/s | bf16 | 17.58 GB | ~238.9 | ~955.5 q4 |
| AMD Instinct MI300X | 192.0 GB | 5300.0 GB/s | bf16 | 17.58 GB | ~211.0 | ~844.0 q4 |
| NVIDIA H200 SXM | 141.0 GB | 4800.0 GB/s | bf16 | 17.58 GB | ~191.1 | ~764.4 q4 |
| NVIDIA H100 NVL | 94.0 GB | 3900.0 GB/s | bf16 | 17.58 GB | ~155.3 | ~621.0 q4 |
| NVIDIA H100 SXM | 80.0 GB | 3350.0 GB/s | bf16 | 17.58 GB | ~133.4 | ~533.5 q4 |
| AMD Instinct MI250X | 128.0 GB | 3200.0 GB/s | bf16 | 17.58 GB | ~127.4 | ~509.6 q4 |
| Google TPU v5p | 95.0 GB | 2765.0 GB/s | bf16 | 17.58 GB | ~110.1 | ~440.3 q4 |
| NVIDIA A100 80GB SXM | 80.0 GB | 2039.0 GB/s | bf16 | 17.58 GB | ~81.2 | ~324.7 q4 |
| NVIDIA H100 PCIe | 80.0 GB | 2000.0 GB/s | bf16 | 17.58 GB | ~79.6 | ~318.5 q4 |
| NVIDIA GeForce RTX 5090 | 32.0 GB | 1792.0 GB/s | bf16 | 17.58 GB | ~71.3 | ~285.4 q4 |
| Google TPU v6e (Trillium) | 32.0 GB | 1638.0 GB/s | bf16 | 17.58 GB | ~65.2 | ~260.8 q4 |
| AMD Instinct MI210 | 64.0 GB | 1600.0 GB/s | bf16 | 17.58 GB | ~63.7 | ~254.8 q4 |
| NVIDIA A100 40GB SXM | 40.0 GB | 1555.0 GB/s | bf16 | 17.58 GB | ~61.9 | ~247.6 q4 |
| Google TPU v4 | 32.0 GB | 1200.0 GB/s | bf16 | 17.58 GB | ~47.8 | ~191.1 q4 |
| NVIDIA GeForce RTX 4090 | 24.0 GB | 1008.0 GB/s | bf16 | 17.58 GB | ~40.1 | ~160.5 q4 |
| AMD Radeon RX 7900 XTX | 24.0 GB | 960.0 GB/s | bf16 | 17.58 GB | ~38.2 | ~152.9 q4 |
| NVIDIA GeForce RTX 5080 | 16.0 GB | 960.0 GB/s | fp8 | 8.79 GB | ~76.4 | ~152.9 q4 |
| NVIDIA GeForce RTX 3090 | 24.0 GB | 936.0 GB/s | bf16 | 17.58 GB | ~37.3 | ~149.1 q4 |
| NVIDIA L40S | 48.0 GB | 864.0 GB/s | bf16 | 17.58 GB | ~34.4 | ~137.6 q4 |
| Apple M3 Ultra | 512.0 GB | 819.0 GB/s | bf16 | 17.58 GB | ~32.6 | ~130.4 q4 |
| AMD Radeon RX 7900 XT | 20.0 GB | 800.0 GB/s | fp8 | 8.79 GB | ~63.7 | ~127.4 q4 |
| Apple M2 Ultra | 192.0 GB | 800.0 GB/s | bf16 | 17.58 GB | ~31.8 | ~127.4 q4 |
| Google TPU v5e | 16.0 GB | 800.0 GB/s | fp8 | 8.79 GB | ~63.7 | ~127.4 q4 |
| NVIDIA GeForce RTX 4080 SUPER | 16.0 GB | 736.0 GB/s | fp8 | 8.79 GB | ~58.6 | ~117.2 q4 |
| NVIDIA GeForce RTX 4070 Ti SUPER | 16.0 GB | 672.0 GB/s | fp8 | 8.79 GB | ~53.5 | ~107.0 q4 |
| AMD Radeon RX 9070 XT | 16.0 GB | 640.0 GB/s | fp8 | 8.79 GB | ~51.0 | ~101.9 q4 |
| Apple M4 Max | 128.0 GB | 546.0 GB/s | bf16 | 17.58 GB | ~21.7 | ~86.9 q4 |
| Apple M1 Max | 64.0 GB | 400.0 GB/s | bf16 | 17.58 GB | ~15.9 | ~63.7 q4 |
| Apple M2 Max | 96.0 GB | 400.0 GB/s | bf16 | 17.58 GB | ~15.9 | ~63.7 q4 |
| Apple M3 Max | 128.0 GB | 400.0 GB/s | bf16 | 17.58 GB | ~15.9 | ~63.7 q4 |
| NVIDIA GeForce RTX 3060 12GB | 12.0 GB | 360.0 GB/s | fp8 | 8.79 GB | ~28.7 | ~57.3 q4 |
| NVIDIA L4 | 24.0 GB | 300.0 GB/s | bf16 | 17.58 GB | ~11.9 | ~47.8 q4 |
| NVIDIA GeForce RTX 4060 Ti 16GB | 16.0 GB | 288.0 GB/s | fp8 | 8.79 GB | ~22.9 | ~45.9 q4 |
| Apple M4 Pro | 64.0 GB | 273.0 GB/s | bf16 | 17.58 GB | ~10.9 | ~43.5 q4 |
| Apple M4 | 32.0 GB | 120.0 GB/s | bf16 | 17.58 GB | ~4.8 | ~19.1 q4 |
Taken from the card's availability section.
| Platform | Id on that platform |
|---|---|
| Hugging Face | ibm-granite/granite-4.2-8b |
| Benchmark | Evaluated as | Score | Evidence date | Source kind | |
|---|---|---|---|---|---|
| aa_lcr | Granite 4.2 8B | 45.0% | 2026-09-10 evaluated | independent evaluator | source |
| critpt | Granite 4.2 8B | 0.29% | 2026-09-10 evaluated | independent evaluator | source |
| gdpval_aa | Granite 4.2 8B | 7.25% | 2026-09-10 evaluated | independent evaluator | source |
| gpqa_diamond | Granite 4.2 8B | 63.13% | 2026-09-10 evaluated | independent evaluator | source |
| scicode | Granite 4.2 8B | 31.48% | 2026-09-10 evaluated | independent evaluator | source |
This card reports no benchmark scores yet.
| Benchmark | Catalogue standing | Score |
|---|