NVIDIA · Volta · 16 GB VRAM tier
Best AI Models for NVIDIA Tesla V100 PCIe 16GB
Runs 14B models well; 27–31B only just fits.
Volta with Tensor Cores and 900 GB/s HBM2, but capped at 16 GB. Passive server card; needs added cooling.
| # | Model | Speed | llmrun Score | Quant · VRAM |
|---|---|---|---|---|
| 1 | Qwen3.5 9BAlibabaHow to run → Q4_K_M · 6.6 GB · 92.0 t/s | 92.0 tok/s | Q4_K_M · 6.6 GB | |
| 2 | Qwen3.5 4BAlibaba Q4_K_M · 3.4 GB · 179 t/s | 179 tok/s | Q4_K_M · 3.4 GB | |
| 3 | GPT OSS 20BOpenAI Q4_K_M · 13.1 GB · 174 t/s | 174 tok/s | Q4_K_M · 13.1 GB | |
| 4 | Phi 4Microsoft Q4_K_M · 12.5 GB · 61.4 t/s | 61.4 tok/s | Q4_K_M · 12.5 GB | |
| 5 | Gemma 4 12B ITGoogle Q4_K_M · 9.0 GB · 71.1 t/s | 71.1 tok/s | Q4_K_M · 9.0 GB |
| # | Model | Speed | Coding Score | Quant · VRAM |
|---|---|---|---|---|
| 1 | GPT OSS 20BOpenAIHow to run → Q4_K_M · 13.1 GB · 174 t/s | 174 tok/s | Q4_K_M · 13.1 GB | |
| 2 | Q4_K_M · 7.9 GB · 110 t/s | 110 tok/s | Q4_K_M · 7.9 GB | |
| 3 | Qwen3.5 9BAlibaba Q4_K_M · 6.6 GB · 92.0 t/s | 92.0 tok/s | overall | Q4_K_M · 6.6 GB |
| 4 | Gemma 4 12B ITGoogle Q4_K_M · 9.0 GB · 71.1 t/s | 71.1 tok/s | Q4_K_M · 9.0 GB | |
| 5 | Q4_K_M · 6.5 GB · 94.2 t/s | 94.2 tok/s | Q4_K_M · 6.5 GB |
| # | Model | Speed | Reasoning Score | Quant · VRAM |
|---|---|---|---|---|
| 1 | Qwen3.5 9BAlibabaHow to run → Q4_K_M · 6.6 GB · 92.0 t/s | 92.0 tok/s | Q4_K_M · 6.6 GB | |
| 2 | GPT OSS 20BOpenAI Q4_K_M · 13.1 GB · 174 t/s | 174 tok/s | Q4_K_M · 13.1 GB | |
| 3 | Qwen3 8BAlibaba Q4_K_M · 7.6 GB · 106 t/s | 106 tok/s | Q4_K_M · 7.6 GB | |
| 4 | Q4_K_M · 7.9 GB · 110 t/s | 110 tok/s | Q4_K_M · 7.9 GB | |
| 5 | Gemma 4 12B ITGoogle Q4_K_M · 9.0 GB · 71.1 t/s | 71.1 tok/s | Q4_K_M · 9.0 GB |
| # | Model | Speed | Long context Score | Quant · VRAM |
|---|---|---|---|---|
| 1 | Qwen3.5 9BAlibabaHow to run → Q4_K_M · 6.6 GB · 92.0 t/s | 92.0 tok/s | Q4_K_M · 6.6 GB | |
| 2 | Qwen3.5 4BAlibaba Q4_K_M · 3.4 GB · 179 t/s | 179 tok/s | Q4_K_M · 3.4 GB | |
| 3 | Mistral Nemo Instruct 2407Mistral AI Q4_K_M · 11.0 GB · 72.5 t/s | 72.5 tok/s | Q4_K_M · 11.0 GB | |
| 4 | Q4_K_M · 7.9 GB · 110 t/s | 110 tok/s | Q4_K_M · 7.9 GB | |
| 5 | Gemma 4 12B ITGoogle Q4_K_M · 9.0 GB · 71.1 t/s | 71.1 tok/s | Q4_K_M · 9.0 GB |
| # | Model | Speed ↓ | llmrun Score | Quant · VRAM |
|---|---|---|---|---|
| 1 | Qwen3.5 4BAlibabaHow to run → Q4_K_M · 3.4 GB · 179 t/s | 179 tok/s | Q4_K_M · 3.4 GB | |
| 2 | GPT OSS 20BOpenAI Q4_K_M · 13.1 GB · 174 t/s | 174 tok/s | Q4_K_M · 13.1 GB | |
| 3 | Qwen3.5 9BAlibaba Q4_K_M · 6.6 GB · 92.0 t/s | 92.0 tok/s | Q4_K_M · 6.6 GB | |
| 4 | Gemma 4 12B ITGoogle Q4_K_M · 9.0 GB · 71.1 t/s | 71.1 tok/s | Q4_K_M · 9.0 GB | |
| 5 | Phi 4Microsoft Q4_K_M · 12.5 GB · 61.4 t/s | 61.4 tok/s | Q4_K_M · 12.5 GB |
| # | Model | Speed | Vision Score | Quant · VRAM |
|---|---|---|---|---|
| 1 | Qwen3.5 9BAlibabaHow to run → Q4_K_M · 6.6 GB · 92.0 t/s | 92.0 tok/s | Q4_K_M · 6.6 GB | |
| 2 | Qwen3.5 4BAlibaba Q4_K_M · 3.4 GB · 179 t/s | 179 tok/s | Q4_K_M · 3.4 GB | |
| 3 | Gemma 3 12B ITGoogle Q4_K_M · 10.9 GB · 72.8 t/s | 72.8 tok/s | Q4_K_M · 10.9 GB | |
| 4 | Gemma 3 4B ITGoogle Q4_K_M · 5.0 GB · 206 t/s | 206 tok/s | Q4_K_M · 5.0 GB | |
| 5 | Q4_K_M · 6.5 GB · 94.2 t/s | 94.2 tok/s | Q4_K_M · 6.5 GB |
The Tesla V100 PCIe 16GB runs 14B models well. Volta generation, released 2017.
Tesla V100 PCIe 16GB strengths and weaknesses
Strengths
- Runs 35B MoE models: Qwen3.6 35B A3B at 174 tok/s in 15.7 GB
- Reads long prompts fast: A 4K-token prompt on Qwen3.5 9B starts answering in 1.6 s
Weaknesses
- 70B models don't fit: Llama 3.3 70B Instruct needs 51.3 GB at Q4 vs 16 GB here
- No FP8 or NVFP4: Volta predates both; use GGUF, AWQ or GPTQ builds
Output speed depends on memory bandwidth (900 GB/s). Grey rows need more than 16 GB.
- 9BQwen3.5 9B · Q4_K_M · 6.6 GB92.0tok/s
- 14BQwen3 14B · Q4_K_M · 11.9 GB61.6tok/s
- 39.0tok/s
- 174tok/s
- 70BLlama 3.3 70B Instruct · not at 4-bitneeds 51.3 GB
- 120B MoEGPT OSS 120B · not at 4-bitneeds 70.8 GB
Estimated speeds: Q4_K_M, 4K context, one user. How we estimate
All 1057 models on the Tesla V100 PCIe 16GB
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 |
|---|---|---|---|---|---|---|
Stories15M MOEChat | Q4_K_M | 0.3 GB | 1275.7 t/s | 0K | S100, Runs great | |
NuExtract 1.5Chat | Q4_K_M | 9.0 GB | 172.1 t/s | 131K | S98, Runs great | |
Typhoon Ocr 3BVision | Q4_K_M | 3.2 GB | 222.4 t/s | 128K | S100, Runs great | |
Qwen3 1.7BChatFunctions | Q4_K_M | 3.4 GB | 334.3 t/s | 41K | S100, Runs great | |
Glm 4 9B ChatChat | Q4_K_M | 8.9 GB | 94.4 t/s | 131K | S90, Runs great | |
Chatglm3 6BChat | Q4_K_M | 5.2 GB | 142.0 t/s | 8K | S96, Runs great | |
| Q4_K_M | 6.7 GB | 122.6 t/s | 33K | S95, Runs great | ||
Qwen2.5 Coder 14B InstructChatCodeFunctions | Q4_K_M | 12.4 GB | 61.2 t/s | 33K | A84, Runs well | |
| Q4_K_M | 7.3 GB | 108.5 t/s | 131K | S93, Runs great | ||
Qwen3.5 0.8B BaseVision | Q4_K_M | 0.9 GB | 672.4 t/s | 262K | S100, Runs great | |
| Q4_K_M | 5.0 GB | 126.6 t/s | 4K | S95, Runs great | ||
Yi Coder 9B ChatChatCode | Q4_K_M | 7.2 GB | 100.9 t/s | 131K | S92, Runs great | |
Qwen3 VL 2B ThinkingVisionFunctions | Q4_K_M | 3.5 GB | 323.2 t/s | 262K | S100, Runs great | |
Gemma 4 26B A4BVision | Q4_K_S | 15.5 GB | 164.9 t/s | 262K | B62, Decent | |
| Q4_K_M | 11.8 GB | 66.3 t/s | 262K | S85, Runs great |
Rent NVIDIA Tesla V100 PCIe 16GB in the cloud
Hourly rate on demand for one Tesla V100 PCIe 16GB.
Vast.aiMarketplace
from$0.12/hr
Cost per million tokens
Output-token cost on a Tesla V100 PCIe 16GB rented at $0.12/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 |
|---|---|---|---|
| GPT OSS 20B | 174 tok/s | $0.19$0.0055 input | |
| Gemma 4 26B A4B IT | 165 tok/s | $0.20$0.0058 input | |
| Qwen3.5 9B | 92 tok/s | $0.36$0.013 input | |
| Gemma 4 12B IT | 71 tok/s | $0.47$0.016 input | |
| Qwen3.8 27B | 39 tok/s | $0.85$0.037 input |
Specs
- Architecture
- Volta
- VRAM
- 16 GB HBM2
- Memory bandwidth
- 900.0 GB/s
- FP16 (dense)
- 112.00 TFLOPS
- CUDA cores
- 5,120
- Tensor cores
- 640
- Compute capability
- 7.0
- TDP
- 250 W
- Released
- 2017-06-21
Tesla V100 PCIe 16GB FAQ
What is the largest model the Tesla V100 PCIe 16GB can run?
Gemma 4 26B A4B IT (25.8B) is the largest at 4-bit or better: at Q4_K_S it uses 15.5 GB of 16 GB, and runs at about 165 tok/s. At 3-bit, a bigger model fits too: Gemma 2 27B IT (27.2B) at IQ3_M, 15.0 GB.
Can the Tesla V100 PCIe 16GB run Qwen3.5 9B?
Yes. Qwen3.5 9B fits at Q4_K_M, using 6.6 GB of 16 GB, and it is the best model that runs well on this card.
Can the Tesla V100 PCIe 16GB run 70B models?
No. Llama 3.3 70B Instruct needs 51.3 GB at Q4, more than the 16 GB on this card.
How many tokens per second does the Tesla V100 PCIe 16GB give?
About 92.0 tok/s on Qwen3.5 9B at Q4_K_M, with the first token after 1.6 s on a 4K-token prompt.
What power supply does the Tesla V100 PCIe 16GB need?
Plan a 500 W PSU (2× the 250 W TDP).
Is 16 GB enough for coding and agents?
Yes for coding: GPT OSS 20B, the best of the 2 coding models that fit, runs at about 174 tok/s. For agents, not yet: no agent-scored model fits in less than 96 GB.