AudioMIT
Voxtral Small 24B
mistralai/Voxtral-Small-24B-2507
Parameters
24B
Context
32K
Downloads
245.8K
Likes
519
Architecture
transformer
GPU Requirements
VRAM / RAM / disk estimates by quantization and usage scenario
| Quantization | Minimum | Recommended | Production |
|---|---|---|---|
| FP1616bit | 110.9 GB RAM 166.4 · Disk 108.7 | 116.6 GB RAM 174.8 · Disk 134.7 | 122.8 GB RAM 184.2 · Disk 173.7 |
| FP88bit | 56.8 GB RAM 85.3 · Disk 73.9 | 60.0 GB RAM 90.0 · Disk 99.9 | 63.8 GB RAM 95.7 · Disk 138.9 |
| INT88bit | 56.8 GB RAM 85.3 · Disk 73.9 | 60.0 GB RAM 90.0 · Disk 99.9 | 63.8 GB RAM 95.7 · Disk 138.9 |
| AWQ4bit | 30.8 GB RAM 46.2 · Disk 57.1 | 32.8 GB RAM 49.2 · Disk 83.1 | 35.4 GB RAM 53.1 · Disk 122.1 |
| GPTQ4bit | 31.2 GB RAM 46.7 · Disk 57.3 | 33.1 GB RAM 49.7 · Disk 83.3 | 35.8 GB RAM 53.7 · Disk 122.3 |
| GGUF4bit | 31.8 GB RAM 47.7 · Disk 57.7 | 33.9 GB RAM 50.8 · Disk 83.7 | 36.5 GB RAM 54.8 · Disk 122.7 |
GPU Recommendations
Best ValueBEST
1× H100 SXM
80 GB VRAM
$1,170/mo
≈ RTF 0.11
Effective production throughput ≈ 0.1 RTF
score 0.5988
Cheapest
1× RTX A6000
48 GB VRAM
$226/mo
≈ RTF 0.46
Effective production throughput ≈ 0.3 RTF
score 0.6729
Performance
8× H200 SXM
1128 GB VRAM
$20,440/mo
≈ RTF 0.02
Effective production throughput ≈ 0 RTF
score 0.2197
Min Complexity
1× L40S
48 GB VRAM
$341/mo
≈ RTF 0.41
Effective production throughput ≈ 0.3 RTF
score 0.6712
💬 Community — real deployment experience
0 Articles · 0 Benchmarks
GPUs that fit (single card)
RTX A6000
48 GB VRAM
$226
/mo · vast
L40S
48 GB VRAM
$341
/mo · vast
RTX PRO 6000 WS
48 GB VRAM
$482
/mo · vast
A100 80GB
80 GB VRAM
$876
/mo · vast
RTX PRO 6000 S
48 GB VRAM
$925
/mo · vast