NVIDIA · 8x B200 SXM · Server
Best AI Models for NVIDIA DGX B200 (1.4 TB)
Runs models up to 1.7T parameters, like DeepSeek V4 Pro 0813.
1.4 TB HBM64 TB/s18 PFLOPS FP1614.3 kW2024
| # | Model | Speed | llmrun Score | Quant · Memory |
|---|---|---|---|---|
| 1 | Inkling 2 SmallThinking MachinesHow to run → Q4_K_M · 359.1 GB · 276 t/s | 276 tok/s | Q4_K_M · 359.1 GB | |
| 2 | DeepSeek V4.1 Flash 1017DeepSeek Q4_K_M · 458.4 GB · 376 t/s | 376 tok/s | Q4_K_M · 458.4 GB | |
| 3 | MiMo V2.6 Pro RLXiaomi Q4_K_M · 614.1 GB · 198 t/s | 198 tok/s | Q4_K_M · 614.1 GB | |
| 4 | MiMo V2.6 Flash RLXiaomi Q4_K_M · 186.5 GB · 285 t/s | 285 tok/s | Q4_K_M · 186.5 GB | |
| 5 | Hy3Tencent Q4_K_M · 182.3 GB · 147 t/s | 147 tok/s | Q4_K_M · 182.3 GB |
| # | Model | Speed | Coding Score | Quant · Memory |
|---|---|---|---|---|
| 1 | Q4_K_M · 458.4 GB · 376 t/s | 376 tok/s | Q4_K_M · 458.4 GB | |
| 2 | GLM 5.3Z.ai Q4_K_M · 453.8 GB · 182 t/s | 182 tok/s | Q4_K_M · 453.8 GB | |
| 3 | Q4_K_M · 551.9 GB · 252 t/s | 252 tok/s | Q4_K_M · 551.9 GB | |
| 4 | DeepSeek V4 Pro 0813DeepSeek Q4_K_M · 990.8 GB · 207 t/s | 207 tok/s | Q4_K_M · 990.8 GB | |
| 5 | Hy3Tencent Q4_K_M · 182.3 GB · 147 t/s | 147 tok/s | Q4_K_M · 182.3 GB |
| # | Model | Speed | Reasoning Score | Quant · Memory |
|---|---|---|---|---|
| 1 | Q4_K_M · 458.4 GB · 376 t/s | 376 tok/s | Q4_K_M · 458.4 GB | |
| 2 | Inkling 2 SmallThinking Machines Q4_K_M · 359.1 GB · 276 t/s | 276 tok/s | Q4_K_M · 359.1 GB | |
| 3 | DeepSeek V4.1 FlashDeepSeek Q4_K_M · 458.4 GB · 376 t/s | 376 tok/s | Q4_K_M · 458.4 GB | |
| 4 | GLM 5.3Z.ai Q4_K_M · 453.8 GB · 182 t/s | 182 tok/s | Q4_K_M · 453.8 GB | |
| 5 | Hy3Tencent Q4_K_M · 182.3 GB · 147 t/s | 147 tok/s | Q4_K_M · 182.3 GB |
| # | Model | Speed | Math Score | Quant · Memory |
|---|---|---|---|---|
| 1 | Q4_K_M · 458.4 GB · 376 t/s | 376 tok/s | Q4_K_M · 458.4 GB | |
| 2 | DeepSeek V4.1 FlashDeepSeek Q4_K_M · 458.4 GB · 376 t/s | 376 tok/s | Q4_K_M · 458.4 GB | |
| 3 | GLM 5.3Z.ai Q4_K_M · 453.8 GB · 182 t/s | 182 tok/s | Q4_K_M · 453.8 GB | |
| 4 | Q4_K_M · 551.9 GB · 252 t/s | 252 tok/s | Q4_K_M · 551.9 GB | |
| 5 | Hy3Tencent Q4_K_M · 182.3 GB · 147 t/s | 147 tok/s | Q4_K_M · 182.3 GB |
| # | Model | Speed ↓ | llmrun Score | Quant · Memory |
|---|---|---|---|---|
| 1 | Qwen3.5 4BAlibabaHow to run → Q4_K_M · 3.4 GB · 1595 t/s | 1595 tok/s | Q4_K_M · 3.4 GB | |
| 2 | Qwen3.5 9BAlibaba Q4_K_M · 6.6 GB · 818 t/s | 818 tok/s | Q4_K_M · 6.6 GB | |
| 3 | GPT OSS 20BOpenAI Q4_K_M · 13.1 GB · 550 t/s | 550 tok/s | Q4_K_M · 13.1 GB | |
| 4 | Gemma 4 26B A4B ITGoogle Q4_K_M · 16.5 GB · 459 t/s | 459 tok/s | Q4_K_M · 16.5 GB | |
| 5 | Hy3Tencent Q4_K_M · 182.3 GB · 147 t/s | 147 tok/s | Q4_K_M · 182.3 GB |
| # | Model | Speed | Vision Score | Quant · Memory |
|---|---|---|---|---|
| 1 | Inkling 2 SmallThinking MachinesHow to run → Q4_K_M · 359.1 GB · 276 t/s | 276 tok/s | Q4_K_M · 359.1 GB | |
| 2 | DeepSeek V4.1 Flash 1017DeepSeek Q4_K_M · 458.4 GB · 376 t/s | 376 tok/s | Q4_K_M · 458.4 GB | |
| 3 | DeepSeek V4.1 FlashDeepSeek Q4_K_M · 458.4 GB · 376 t/s | 376 tok/s | Q4_K_M · 458.4 GB | |
| 4 | Q4_K_M · 551.9 GB · 252 t/s | 252 tok/s | Q4_K_M · 551.9 GB | |
| 5 | Muse Glimmer 30BMeta Q4_K_M · 19.6 GB · 284 t/s | 284 tok/s | Q4_K_M · 19.6 GB |
| # | Model | Speed | Agents Score | Quant · Memory |
|---|---|---|---|---|
| 1 | Q4_K_M · 458.4 GB · 376 t/s | 376 tok/s | Q4_K_M · 458.4 GB | |
| 2 | GLM 5.3Z.ai Q4_K_M · 453.8 GB · 182 t/s | 182 tok/s | Q4_K_M · 453.8 GB | |
| 3 | MiMo V2.6 Pro RLXiaomi Q4_K_M · 614.1 GB · 198 t/s | 198 tok/s | Q4_K_M · 614.1 GB | |
| 4 | GLM 5.2Z.ai Q4_K_M · 453.8 GB · 182 t/s | 182 tok/s | Q4_K_M · 453.8 GB | |
| 5 | Hy3Tencent Q4_K_M · 182.3 GB · 147 t/s | 147 tok/s | Q4_K_M · 182.3 GB |
See all 453 models on the NVIDIA DGX B200 →Bars = traction for models without a Score yet
The NVIDIA DGX B200 runs models up to 1.7T parameters, like DeepSeek V4 Pro 0813. A 4K-token prompt waits 80 ms before the first word on Inkling 2 Small. 8x B200 SXM (Blackwell), released 2024.
For models1.4 TBall of it
Chip8x B200 SXMreleased 2024
NVIDIA DGX B200 strengths and weaknesses
Strengths
- Runs 120B MoE models: GPT OSS 120B at 378 tok/s in 70.8 GB
- Room for long context: Llama 3.3 70B Instruct at Q4 uses 51.3 GB, leaving 1.4 TB free
- Reads long prompts fast: A 32K-token prompt on Inkling 2 Small starts answering in 660 ms
- FP8 and NVFP4 run natively: Blackwell takes FP8 and NVFP4 checkpoints in vLLM as published
Memory for models
1.4 TB of 1.4 TB120B MoE70.8 GB1650B990.8 GB
120B MoE70.8 GB1650B990.8 GB
For models 1.4 TBModel size at Q4
Everything on the size ladder fits. Inkling 2 Small and DeepSeek V4.1 Flash 1017 together use 817.5 GB and leave 622.5 GB free.
The same chip from other makers first, then machines and GPUs people weigh against it, each with the one reason it is worth a look. Speeds on Qwen3.6 35B A3B.
Step down to 508.5 GBMac Studio M5 Ultra (512 GB)Apple · 508.5 GB for models · M5 Ultra
▼ 76% speed▼ 931.5 GBabout $22,299
Step down to 508.5 GBMac Studio M3 Ultra (512 GB)Apple · 508.5 GB for models · M3 Ultra▼ 79% speed▼ 931.5 GBabout $9,499
Step down to 188.5 GBMac Studio M2 Ultra (192 GB)Apple · 188.5 GB for models · M2 Ultra▼ 79% speed▼ 1.3 TBabout $5,599
Step down to 124.5 GBMac Studio M4 Max (128 GB)Apple · 124.5 GB for models · M4 Max▼ 82% speed▼ 1.3 TBabout $3,499
Step down to 24 GBRTX 4090NVIDIA · 24 GB VRAM · GPU▼ 58% speed▼ 1.4 TBabout $1,599rent $0.34/hr
Step down to 32 GBRTX 5090NVIDIA · 32 GB VRAM · GPU▼ 41% speed▼ 1.4 TBabout $5,502rent $0.50/hr
Output speed depends on memory bandwidth: each model runs on as few of the 8 GPUs as hold it, at 8 TB/s per GPU. Grey rows need more than the 1.4 TB available.
- 9BQwen3.5 9B · Q4_K_M · 6.6 GB818tok/s
- 14BQwen3 14B · Q4_K_M · 11.9 GB547tok/s
- 27–31BQwen3.8 27B · Q4_K_M · 17.9 GB299tok/s
- 35B MoEQwen3.6 35B A3B · Q4_K_M · 22.0 GB363tok/s
- 70BLlama 3.3 70B Instruct · Q4_K_M · 51.3 GB112tok/s
- 120B MoEGPT OSS 120B · Q4_K_M · 70.8 GB378tok/s
Estimated speeds: Q4_K_M, 4K context, one user. How we estimate
All 1390 models on the NVIDIA DGX B200
Featured onlyBest quant that fits in 1440 GB, ordered by llmrun Score. Fit grades how well it runs here, S best. How ranking works
| Model | Quant | VRAM | Speed | Context | Score | Fit |
|---|---|---|---|---|---|---|
| Q4_K_M | 7.3 GB | 964.7 t/s | 131K | S100, Runs great | ||
Jan v3 4B Base InstructChatCode | Q4_K_M | 4.5 GB | 1656.1 t/s | 262K | S100, Runs great | |
| BF16 | 7.6 GB | 862.4 t/s | 29K | S100, Runs great | ||
Olmo 3 7B InstructChatFunctions | Q4_K_M | 8.4 GB | 904.3 t/s | 66K | S100, Runs great | |
| Q4_K_M | 5.8 GB | 1042.1 t/s | 33K | S100, Runs great | ||
Aya Expanse 8BChat | BF16 | 20.2 GB | 294.5 t/s | — | S100, Runs great | |
Qwen2.5 3B InstructChatFunctions | Q4_K_M | 2.8 GB | 2331.8 t/s | 33K | S100, Runs great | |
| Q4_K_M | 2.3 GB | 3113.8 t/s | 4K | S100, Runs great | ||
WizardLM 2 8x22BChat | Q4_K_M | 88.4 GB | 130.9 t/s | 66K | S95, Runs great | |
Granite Docling 258MVisionChatCodeMath | Q4_K_M | 0.6 GB | 10400.0 t/s | 8K | S100, Runs great | |
Tiny Dummy Qwen2Chat | Q4_K_M | 0.5 GB | 17333.3 t/s | 33K | S100, Runs great | |
PowerMoE 3BChat | Q4_K_M | 2.6 GB | 527.9 t/s | 4K | S100, Runs great | |
| Q4_K_M | 4.1 GB | 2113.8 t/s | 131K | S100, Runs great | ||
NousCoder 14BChatCode | Q4_K_M | 11.8 GB | 547.4 t/s | 82K | S100, Runs great | |
| Q4_K_M | 17.8 GB | 349.9 t/s | 33K | S100, Runs great |
Specs
- Chip
- 8x B200 SXM · NVIDIA B200
- CPU cores
- 112
- Memory
- 1.4 TB
- For models
- 1.4 TB
- Memory bandwidth
- 64 TB/s
- FP16 (dense)
- 18 PFLOPS
- Power
- 14.3 kW max
- Size
- 10U air-cooled
- Released
- 2024-10-08
NVIDIA DGX B200 FAQ
How much memory is available for AI on the NVIDIA DGX B200?
All 1.4 TB can hold model weights.
What is the largest model the NVIDIA DGX B200 can run?
DeepSeek V4 Pro 0813 (1650.5B) is the largest at 4-bit or better: at Q4_K_M it uses 990.8 GB, 69% of the available memory, and runs at about 207 tok/s. At 3-bit, a bigger model fits too: Kimi K3 (2779.9B) at Q3_K_M, 1357.3 GB.
Can the NVIDIA DGX B200 run Inkling 2 Small?
Yes. Inkling 2 Small fits at Q4_K_M, using 359.1 GB of the 1.4 TB available, and it is the best model that runs well on this machine.
How long does the NVIDIA DGX B200 take to read a long prompt?
On Inkling 2 Small, about 80 ms for a 4K-token prompt and 660 ms for a 32K-token document before the first word appears.