LG AI Research · lgai-exaone/exaone-4-0-32b
| Provider | LG AI Research |
|---|---|
| Family | exaone |
| Type | llm-chat |
| Status | active |
| Released | 2025-07-11 |
| Updated | 2025-08-04 |
| Parameters | 32B |
| Open weights | yes |
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 | fp8 | 32.0 GB | ~459328.8 | ~918657.7 q4 |
| NVIDIA GB200 Grace Blackwell Superchip | 372.0 GB | 16000.0 GB/s | bf16 | 64.01 GB | ~175.0 | ~699.9 q4 |
| AMD Instinct MI355X | 288.0 GB | 8000.0 GB/s | bf16 | 64.01 GB | ~87.5 | ~350.0 q4 |
| NVIDIA B200 | 180.0 GB | 7700.0 GB/s | bf16 | 64.01 GB | ~84.2 | ~336.8 q4 |
| Google TPU7x (Ironwood) | 192.0 GB | 7380.0 GB/s | bf16 | 64.01 GB | ~80.7 | ~322.8 q4 |
| AMD Instinct MI325X | 256.0 GB | 6000.0 GB/s | bf16 | 64.01 GB | ~65.6 | ~262.5 q4 |
| AMD Instinct MI300X | 192.0 GB | 5300.0 GB/s | bf16 | 64.01 GB | ~58.0 | ~231.9 q4 |
| NVIDIA H200 SXM | 141.0 GB | 4800.0 GB/s | bf16 | 64.01 GB | ~52.5 | ~210.0 q4 |
| NVIDIA H100 NVL | 94.0 GB | 3900.0 GB/s | bf16 | 64.01 GB | ~42.7 | ~170.6 q4 |
| NVIDIA H100 SXM | 80.0 GB | 3350.0 GB/s | fp8 | 32.0 GB | ~73.3 | ~146.5 q4 |
| AMD Instinct MI250X | 128.0 GB | 3200.0 GB/s | bf16 | 64.01 GB | ~35.0 | ~140.0 q4 |
| Google TPU v5p | 95.0 GB | 2765.0 GB/s | bf16 | 64.01 GB | ~30.2 | ~121.0 q4 |
| NVIDIA A100 80GB SXM | 80.0 GB | 2039.0 GB/s | fp8 | 32.0 GB | ~44.6 | ~89.2 q4 |
| NVIDIA H100 PCIe | 80.0 GB | 2000.0 GB/s | fp8 | 32.0 GB | ~43.7 | ~87.5 q4 |
| NVIDIA GeForce RTX 5090 | 32.0 GB | 1792.0 GB/s | q5 | 20.0 GB | ~62.7 | ~78.4 q4 |
| Google TPU v6e (Trillium) | 32.0 GB | 1638.0 GB/s | q5 | 20.0 GB | ~57.3 | ~71.7 q4 |
| AMD Instinct MI210 | 64.0 GB | 1600.0 GB/s | fp8 | 32.0 GB | ~35.0 | ~70.0 q4 |
| NVIDIA A100 40GB SXM | 40.0 GB | 1555.0 GB/s | q6 | 24.0 GB | ~45.3 | ~68.0 q4 |
| Google TPU v4 | 32.0 GB | 1200.0 GB/s | q5 | 20.0 GB | ~42.0 | ~52.5 q4 |
| NVIDIA GeForce RTX 4090 | 24.0 GB | 1008.0 GB/s | q4 | 16.0 GB | ~44.1 | ~44.1 q4 |
| AMD Radeon RX 7900 XTX | 24.0 GB | 960.0 GB/s | q4 | 16.0 GB | ~42.0 | ~42.0 q4 |
| NVIDIA GeForce RTX 3090 | 24.0 GB | 936.0 GB/s | q4 | 16.0 GB | ~40.9 | ~40.9 q4 |
| NVIDIA L40S | 48.0 GB | 864.0 GB/s | fp8 | 32.0 GB | ~18.9 | ~37.8 q4 |
| Apple M3 Ultra | 512.0 GB | 819.0 GB/s | bf16 | 64.01 GB | ~9.0 | ~35.8 q4 |
| Apple M2 Ultra | 192.0 GB | 800.0 GB/s | bf16 | 64.01 GB | ~8.7 | ~35.0 q4 |
| Apple M4 Max | 128.0 GB | 546.0 GB/s | bf16 | 64.01 GB | ~6.0 | ~23.9 q4 |
| Apple M1 Max | 64.0 GB | 400.0 GB/s | fp8 | 32.0 GB | ~8.7 | ~17.5 q4 |
| Apple M2 Max | 96.0 GB | 400.0 GB/s | bf16 | 64.01 GB | ~4.4 | ~17.5 q4 |
| Apple M3 Max | 128.0 GB | 400.0 GB/s | bf16 | 64.01 GB | ~4.4 | ~17.5 q4 |
| NVIDIA L4 | 24.0 GB | 300.0 GB/s | q4 | 16.0 GB | ~13.1 | ~13.1 q4 |
| Apple M4 Pro | 64.0 GB | 273.0 GB/s | fp8 | 32.0 GB | ~6.0 | ~11.9 q4 |
| Apple M4 | 32.0 GB | 120.0 GB/s | q5 | 20.0 GB | ~4.2 | ~5.2 q4 |
Taken from the card's availability section.
| Platform | Id on that platform |
|---|---|
| Hugging Face | LGAI-EXAONE/EXAONE-4.0-32B |
| Benchmark | Evaluated as | Score | Evidence date | Source kind | |
|---|---|---|---|---|---|
| critpt | EXAONE 4.0 32B | 0.0% | 2026-09-10 evaluated | independent evaluator | source |
| gpqa_diamond | EXAONE 4.0 32B | 73.94% | 2026-09-10 evaluated | independent evaluator | source |
This card reports no benchmark scores yet.
| Benchmark | Catalogue standing | Score |
|---|