B300
Next-generation data center GPU based on NVIDIA Blackwell Ultra architecture with 288 GB HBM3e memory, delivering 1.5x FP4 performance and 2x attention performance over the B200 for AI reasoning and real-time inference workloads.
GPU Benchmarks
B300 vs B200
Compare performance across LLMs and image models to find the best GPU for your workload.
Measured B300 vs B200 performance
Runpod measured both GPUs in September 2026 on single-GPU Secure Cloud pods. LLM benchmarks used vLLM 0.28/0.29, and image benchmarks used Hugging Face Diffusers. These figures replace the earlier B300 estimates.
LLM throughput
| Workload | B300 | B200 |
|---|---|---|
| Llama 3.1 8B, single user | 242.3 tok/s | 258.5 tok/s |
| Llama 3.1 8B, chat peak | 16,855 tok/s | 19,337 tok/s |
| Llama 3.1 8B, RAG peak | 13,624 tok/s | 12,641 tok/s |
| Llama 3.1 8B, long-generation observed peak | 23,582 tok/s* | 22,823 tok/s |
| Qwen2.5 7B, single user | 266.1 tok/s | 273.7 tok/s |
| Qwen2.5 7B, chat peak | 16,769 tok/s | 17,713 tok/s |
| Qwen2.5 7B, RAG observed peak | 17,157 tok/s | 15,998 tok/s* |
| Qwen2.5 7B, long-generation observed peak | 24,411 tok/s | 14,445 tok/s* |
*The request-rate ladder had not reached a verified plateau, so the marked result is a measured lower bound rather than a confirmed ceiling. Single-user rows use one request in flight. Peak rows report the highest sustained output rate observed on the workload ladder.
Image generation
| Workload | B300 | B200 |
|---|---|---|
| FLUX.1-dev, 1024x1024, 28 steps, batch 1 | 4.01 sec/image | 3.25 sec/image |
| FLUX.1-dev, 1024x1024, 28 steps, batch 4 | 3.84 sec/image | 3.20 sec/image |
| SDXL, 1024x1024, 20 steps, batch 1 | 0.90 sec/image | Not measured |
| SDXL, 1024x1024, 40 steps, batch 1 | 1.67 sec/image | Not measured |
| SDXL, 1024x1024, 20 steps, batch 4 | 0.57 sec/image | Not measured |
B200 SDXL data was still pending because the pinned datacenter runs had not secured a healthy B200 host. For FLUX.1-dev, lower time is better, and B200 was faster in this benchmark vintage.
What the measurements show
For these 7B and 8B bf16 models, the two Blackwell GPUs are close at single-user speed and most Llama workloads. B200 leads on both chat peaks and FLUX.1-dev. B300 leads on the measured Llama RAG and long-generation rows and the measured Qwen RAG and long-generation rows, subject to the lower-bound flags above.
Method and caveats
- Source vintage: Runpod Chirp, pulled September 15, 2026 at 20:15 UTC.
- LLM ceiling measurements vary 10% to 30% between hosts of the same GPU. Single-user measurements varied 1% to 5%.
- Image comparisons use 1024x1024. On fast cards, 512x512 results are sensitive to CPU and kernel-launch overhead.
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.
LLM inference benchmarks.
Runpod benchmark measurements from September 2026, using vLLM.
B200
Next-generation data center GPU based on Blackwell architecture that features 180 GB of HBM3e memory with 8TB/s bandwidth, delivering up to 20 petaFLOPS of FP4 AI compute performance.
H100 PCIe
High-efficiency LLM processing at 90.98 tok/s.
Image generation benchmarks.
Runpod benchmark measurements from September 2026, using Hugging Face Diffusers.
H100 SXM
Unmatched image gen speed with 49.9 images per minute.
B200
AI image processing at 40.3 images per minute.
H100 PCIe
Pro-grade performance with 36 images per minute.
Related Comparisons
- B300 vs L4
- B300 vs H100 SXM
- B300 vs H200
- B300 vs RTX 3090
- B300 vs RTX A6000
- B300 vs A40
- B300 vs H100 NVL
- B300 vs L40
- B300 vs H100 PCIe
Case Studies
Real-world GPU performance in action.
See how teams optimize cost and performance with the right GPU for their workloads.
"We didn't seriously evaluate other alternatives before committing to Runpod. We found their approach interesting from the beginning, particularly the ease of integration, and we haven't had a reason to switch since."
.png){kind=small}
"AWS is very reliable and has everything we need, but it is way more expensive than Runpod. With Runpod, we can choose the right card for the job."
.png){kind=logo}
"Runpod is easily one of the most scalable tools that we use in our infrastructure."
.png){kind=small}
"What's happened is, moving to Runpod, Botto has a better GPU, better graphic models and better resources, but it also costs less. That's something I didn't expect we were going to have."
"You guys have genuinely changed the scope of what we've been able to attempt."
"All of these projects, the renders for AMD, the Coca-Cola builds, that has to do with scalability. If we can't scale, we can't deliver. Runpod makes that possible."
"Runpod has changed the way we ship because we no longer have to wonder if we have access to GPUs. We've saved probably 90% on our infrastructure bill, mainly because we can use bursty compute whenever we need it."
"Runpod has allowed the team to focus more on the features that are core to our product and that are within our skill set, rather than spending time focusing on infrastructure, which can sometimes be a bit of a distraction.”
"Runpod helped us scale the part of our platform that drives creation. That’s what fuels the rest, image generation, sharing, remixing. It starts with training."
"Runpod allowed us to reliably handle scaling from zero to over 1,000 requests per second in our live application."
"Runpod has allowed us to focus entirely on growth and product development without us having to worry about the GPU infrastructure at all."
"We could stop worrying about infrastructure and go back to building. That’s the real win.”
“The main value proposition for us was the flexibility Runpod offered. We were able to scale up effortlessly to meet the demand at launch.”
"After migration, we were able to cut down our server costs from thousands of dollars per day to only hundreds."
Runpod’s scalable GPU infrastructure gave us the flexibility we needed to match customer traffic and model complexity, without overpaying for idle resources.
