NVIDIA · Blackwell
Best AI Models for NVIDIA B200 (192 GB)
Runs everything up to 120B MoE well.
192 GB HBM3e8000 GB/s2250 TFLOPS FP161000 WBlackwell 2024
Datacenter accelerator — sold in HGX/NVL systems and typically rented in the cloud, not installed in a desktop. Listed for comparison.
| # | Model | Speed | llmrun Score | Quant · VRAM |
|---|---|---|---|---|
| 1 | Qwen3.8 27BAlibabaHow to run → Q4_K_M · 17.9 GB · 299 t/s | 299 tok/s | Q4_K_M · 17.9 GB | |
| 2 | Inkling SmallThinking Machines Q4_K_M · 162.7 GB · 399 t/s | 399 tok/s | Q4_K_M · 162.7 GB | |
| 3 | DeepSeek V4 FlashDeepSeek Q4_K_M · 175.1 GB · 250 t/s | 250 tok/s | Q4_K_M · 175.1 GB | |
| 4 | Qwen3.6 27BAlibaba Q4_K_M · 17.9 GB · 299 t/s | 299 tok/s | Q4_K_M · 17.9 GB | |
| 5 | Q4_K_M · 19.3 GB · 319 t/s | 319 tok/s | Q4_K_M · 19.3 GB |
| # | Model | Speed | Coding Score | Quant · VRAM |
|---|---|---|---|---|
| 1 | Qwen3.8 27BAlibabaHow to run → Q4_K_M · 17.9 GB · 299 t/s | 299 tok/s | Q4_K_M · 17.9 GB | |
| 2 | DeepSeek V4 FlashDeepSeek Q4_K_M · 175.1 GB · 250 t/s | 250 tok/s | Q4_K_M · 175.1 GB | |
| 3 | Gemma 4 31B ITGoogle Q4_K_M · 21.7 GB · 255 t/s | 255 tok/s | Q4_K_M · 21.7 GB | |
| 4 | Qwen3.6 35B A3BAlibaba Q4_K_M · 22.0 GB · 363 t/s | 363 tok/s | Q4_K_M · 22.0 GB | |
| 5 | Q4_K_M · 19.3 GB · 319 t/s | 319 tok/s | Q4_K_M · 19.3 GB |
| # | Model | Speed | Reasoning Score | Quant · VRAM |
|---|---|---|---|---|
| 1 | Qwen3.8 27BAlibabaHow to run → Q4_K_M · 17.9 GB · 299 t/s | 299 tok/s | Q4_K_M · 17.9 GB | |
| 2 | Inkling SmallThinking Machines Q4_K_M · 162.7 GB · 399 t/s | 399 tok/s | Q4_K_M · 162.7 GB | |
| 3 | Gemma 4 31B ITGoogle Q4_K_M · 21.7 GB · 255 t/s | 255 tok/s | Q4_K_M · 21.7 GB | |
| 4 | DeepSeek V4 FlashDeepSeek Q4_K_M · 175.1 GB · 250 t/s | 250 tok/s | Q4_K_M · 175.1 GB | |
| 5 | Qwen3.8 Flash NextAlibaba Q4_K_M · 109.0 GB · 102 t/s | 102 tok/s | Q4_K_M · 109.0 GB |
| # | Model | Speed | Math Score | Quant · VRAM |
|---|---|---|---|---|
| 1 | Qwen3.8 27BAlibabaHow to run → Q4_K_M · 17.9 GB · 299 t/s | 299 tok/s | Q4_K_M · 17.9 GB | |
| 2 | Inkling SmallThinking Machines Q4_K_M · 162.7 GB · 399 t/s | 399 tok/s | Q4_K_M · 162.7 GB | |
| 3 | Qwen3.6 27BAlibaba Q4_K_M · 17.9 GB · 299 t/s | 299 tok/s | Q4_K_M · 17.9 GB | |
| 4 | Qwen3.6 35B A3BAlibaba Q4_K_M · 22.0 GB · 363 t/s | 363 tok/s | Q4_K_M · 22.0 GB | |
| 5 | Q4_K_M · 19.3 GB · 319 t/s | 319 tok/s | Q4_K_M · 19.3 GB |
| # | Model | Speed | Long context Score | Quant · VRAM |
|---|---|---|---|---|
| 1 | Qwen3.8 27BAlibabaHow to run → Q4_K_M · 17.9 GB · 299 t/s | 299 tok/s | Q4_K_M · 17.9 GB | |
| 2 | Qwen3.6 27BAlibaba Q4_K_M · 17.9 GB · 299 t/s | 299 tok/s | Q4_K_M · 17.9 GB | |
| 3 | MiniMax M2.7MiniMax Q4_K_M · 139.6 GB · 213 t/s | 213 tok/s | Q4_K_M · 139.6 GB | |
| 4 | Gemma 4 31B ITGoogle Q4_K_M · 21.7 GB · 255 t/s | 255 tok/s | Q4_K_M · 21.7 GB | |
| 5 | Q4_K_M · 19.3 GB · 319 t/s | 319 tok/s | Q4_K_M · 19.3 GB |
| # | Model | Speed ↓ | llmrun Score | Quant · VRAM |
|---|---|---|---|---|
| 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 | Q4_K_M · 19.3 GB · 319 t/s | 319 tok/s | Q4_K_M · 19.3 GB |
| # | Model | Speed | Vision Score | Quant · VRAM |
|---|---|---|---|---|
| 1 | Qwen3.8 27BAlibabaHow to run → Q4_K_M · 17.9 GB · 299 t/s | 299 tok/s | Q4_K_M · 17.9 GB | |
| 2 | Inkling SmallThinking Machines Q4_K_M · 162.7 GB · 399 t/s | 399 tok/s | Q4_K_M · 162.7 GB | |
| 3 | Qwen3.6 27BAlibaba Q4_K_M · 17.9 GB · 299 t/s | 299 tok/s | Q4_K_M · 17.9 GB | |
| 4 | Gemma 4 31B ITGoogle Q4_K_M · 21.7 GB · 255 t/s | 255 tok/s | Q4_K_M · 21.7 GB | |
| 5 | Q4_K_M · 87.7 GB · 240 t/s | 240 tok/s | Q4_K_M · 87.7 GB |
See all 419 models on the B200 →Best quant that fits · speeds estimated on this card
The B200 runs everything up to 120B MoE well. Blackwell generation, released 2024.
GenerationBlackwellreleased 2024
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 140.7 GB free
- Reads long prompts fast: A 4K-token prompt on Qwen3.8 27B starts answering in 225 ms
- FP8 and NVFP4 run natively: Blackwell takes FP8 and NVFP4 checkpoints in vLLM as published
Weaknesses
- 1000 W under load: Plan a 2,000 W PSU; ≈$219 a year at 4 h a day
Output speed depends on memory bandwidth (8000 GB/s).
- 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 1318 models on the B200
Featured onlyBest quant that fits, ordered by llmrun Score. Fit grades how well it runs here, S best. How ranking works
| Model | Quant | VRAM | Speed | Context | Score | Fit |
|---|---|---|---|---|---|---|
XCurOS0.1 8B InstructChatFunctions | BF16 | 16.5 GB | 332.3 t/s | 33K | S100, Runs great | |
Ornith 1.0 35BChatFunctions | Q4_K_M | 21.5 GB | 376.2 t/s | 262K | S100, Runs great | |
Qwen3 30B A3B Instruct 2507ChatFunctions | Q4_K_M | 19.4 GB | 329.5 t/s | 262K | S100, Runs great | |
MiniCPM5 1BChatFunctions | Q4_K_M | 1.3 GB | 5252.5 t/s | 131K | S100, Runs great | |
Qwen2.5 7B PyuigptChatFunctions | Q4_K_M | 5.8 GB | 1042.1 t/s | 33K | S100, Runs great | |
Qwythos 9B Claude Mythos 5 1MChatReasoningFunctions | Q4_K_M | 6.5 GB | 837.4 t/s | 1049K | S100, Runs great | |
| Q4_K_M | 0.3 GB | 14444.4 t/s | 1K | S100, Runs great | ||
GLM 5.3 Flash BF16VisionFunctions | Q3_K_M | 169.0 GB | 237.4 t/s | 1049K | S90, Runs great | |
| Q4_K_M | 17.5 GB | 362.6 t/s | 33K | S100, Runs great | ||
Qwen3.6 27B AEON Ultimate Uncensored BF16ChatFunctionsReasoningVision | Q4_K_M | 17.6 GB | 303.0 t/s | 262K | S100, Runs great | |
Nex N2 MiniChat | BF16 | 70.7 GB | 269.6 t/s | 262K | S100, Runs great | |
Pantheon Reasoning 27BChatRoleplayReasoning | Q4_K_M | 17.9 GB | 298.5 t/s | 262K | S100, Runs great | |
LLaDA2.1 MiniChatFunctions | BF16 | 33.5 GB | 603.2 t/s | 33K | S100, Runs great | |
| Q4_K_M | 3.5 GB | 1516.0 t/s | 131K | S100, Runs great | ||
Qwen3 0.6BChatFunctions | Q4_K_M | 1.7 GB | 5977.0 t/s | 41K | S100, Runs great |
Rent NVIDIA B200 in the cloud
Hourly rate on demand for one B200.
Cost per million tokens
Output-token cost on a B200 rented at $5.98/hr, one request at a time. Batch several users on the same card and the cost per token drops several-fold.
| Model | Quant | Speed | Per 1M output |
|---|---|---|---|
| Inkling Small | 399 tok/s | $4.17$0.033 input | |
| Qwen3.8 27B | 299 tok/s | $5.56$0.091 input | |
| Muse Glimmer 30B | 284 tok/s | $5.86$0.098 input | |
| GLM 5.3 Flash | 228 tok/s | $7.29$0.060 input | |
| DeepSeek V4 Flash 0731 | 181 tok/s | $9.18$0.090 input |
Specs
- Architecture
- Blackwell
- VRAM
- 192 GB HBM3e
- Memory bandwidth
- 8000.0 GB/s
- FP16 (dense)
- 2250.00 TFLOPS
- Compute capability
- 10.0
- TDP
- 1000 W
- Released
- 2024-03-18
B200 FAQ
What is the largest model the B200 can run?
MiMo V2.6 Flash RL (309B) is the largest at 4-bit or better: at Q4_K_M it uses 186.5 GB of 192 GB, and runs at about 229 tok/s.
Can the B200 run Qwen3.8 27B?
Yes. Qwen3.8 27B fits at Q4_K_M, using 17.9 GB of 192 GB, and it is the best model that runs well on this card.
How many tokens per second does the B200 give?
About 299 tok/s on Qwen3.8 27B at Q4_K_M, with the first token after 0.2 s on a 4K-token prompt.
What power supply does the B200 need?
Plan a 2000 W PSU (2× the 1000 W TDP).
Is 192 GB enough for coding and agents?
Yes for coding: Qwen3.8 27B, the best of the 19 coding models that fit, runs at about 299 tok/s.