L
localai.computer
ModelsGPUsSystemsAI SetupsBuildsMethodology

Resources

  • Methodology
  • Submit Benchmark
  • About

Browse

  • AI Models
  • GPUs
  • PC Builds

Community

  • Leaderboard

Legal

  • Privacy
  • Terms
  • Contact

© 2025 localai.computer. Hardware recommendations for running AI models locally.

ℹ️We earn from qualifying purchases through affiliate links at no extra cost to you. This supports our free content and research.

  1. Home
  2. GPUs
  3. NVIDIA A100 40GB PCIe

Quick Answer: NVIDIA A100 40GB PCIe offers 40GB VRAM and starts around current market pricing. It delivers approximately 305 tokens/sec on deepseek-ai/DeepSeek-OCR. It typically draws 250W under load.

NVIDIA A100 40GB PCIe

Check availability
By NVIDIAReleased 2020-05MSRP $9,000.00

This GPU offers reliable throughput for local AI workloads. Pair it with the right model quantization to hit your desired tokens/sec, and monitor prices below to catch the best deal.

Search on AmazonView Benchmarks
Specs snapshot
Key hardware metrics for AI workloads.
VRAM40GB
Cores6,912
TDP250W
ArchitectureAmpere

Where to Buy

Buy directly on Amazon with fast shipping and reliable customer service.

No purchase links available yet. Try the Amazon search results to find this GPU.

💡 Not ready to buy? Try cloud GPUs first

Test NVIDIA A100 40GB PCIe performance in the cloud before investing in hardware. Pay by the hour with no commitment.

Vast.aifrom $0.20/hrRunPodfrom $0.30/hrLambda Labsenterprise-grade

AI benchmarks

ModelQuantizationTokens/secVRAM used
deepseek-ai/DeepSeek-OCRQ4
305.06 tok/sEstimated

Auto-generated benchmark

2GB
google/gemma-2bQ4
299.56 tok/sEstimated

Auto-generated benchmark

1GB
google-t5/t5-3bQ4
292.06 tok/sEstimated

Auto-generated benchmark

2GB
unsloth/Llama-3.2-1B-InstructQ4
292.02 tok/sEstimated

Auto-generated benchmark

1GB
meta-llama/Llama-3.2-1BQ4
291.66 tok/sEstimated

Auto-generated benchmark

1GB
nari-labs/Dia2-2BQ4
291.41 tok/sEstimated

Auto-generated benchmark

2GB
unsloth/Llama-3.2-3B-InstructQ4
288.71 tok/sEstimated

Auto-generated benchmark

2GB
meta-llama/Llama-3.2-3B-InstructQ4
288.46 tok/sEstimated

Auto-generated benchmark

2GB
ibm-research/PowerMoE-3bQ4
286.95 tok/sEstimated

Auto-generated benchmark

2GB
bigcode/starcoder2-3bQ4
286.26 tok/sEstimated

Auto-generated benchmark

2GB
google/gemma-3-1b-itQ4
286.00 tok/sEstimated

Auto-generated benchmark

1GB
facebook/sam3Q4
283.67 tok/sEstimated

Auto-generated benchmark

1GB
google-bert/bert-base-uncasedQ4
279.10 tok/sEstimated

Auto-generated benchmark

1GB
inference-net/Schematron-3BQ4
278.91 tok/sEstimated

Auto-generated benchmark

2GB
google/gemma-2-2b-itQ4
278.86 tok/sEstimated

Auto-generated benchmark

1GB
unsloth/gemma-3-1b-itQ4
272.47 tok/sEstimated

Auto-generated benchmark

1GB
meta-llama/Llama-3.2-3BQ4
267.77 tok/sEstimated

Auto-generated benchmark

2GB
WeiboAI/VibeThinker-1.5BQ4
266.87 tok/sEstimated

Auto-generated benchmark

1GB
allenai/OLMo-2-0425-1BQ4
265.74 tok/sEstimated

Auto-generated benchmark

1GB
context-labs/meta-llama-Llama-3.2-3B-Instruct-FP16Q4
263.70 tok/sEstimated

Auto-generated benchmark

2GB
meta-llama/Llama-3.2-1B-InstructQ4
261.73 tok/sEstimated

Auto-generated benchmark

1GB
Qwen/Qwen2.5-3BQ4
261.26 tok/sEstimated

Auto-generated benchmark

2GB
meta-llama/Llama-Guard-3-1BQ4
258.34 tok/sEstimated

Auto-generated benchmark

1GB
deepseek-ai/deepseek-coder-1.3b-instructQ4
257.39 tok/sEstimated

Auto-generated benchmark

2GB
LiquidAI/LFM2-1.2BQ4
255.31 tok/sEstimated

Auto-generated benchmark

1GB
apple/OpenELM-1_1B-InstructQ4
253.25 tok/sEstimated

Auto-generated benchmark

1GB
google/embeddinggemma-300mQ4
253.20 tok/sEstimated

Auto-generated benchmark

1GB
TinyLlama/TinyLlama-1.1B-Chat-v1.0Q4
253.12 tok/sEstimated

Auto-generated benchmark

1GB
ibm-granite/granite-3.3-2b-instructQ4
252.58 tok/sEstimated

Auto-generated benchmark

1GB
IlyaGusev/saiga_llama3_8bQ4
249.78 tok/sEstimated

Auto-generated benchmark

4GB
deepseek-ai/DeepSeek-R1-0528Q4
249.76 tok/sEstimated

Auto-generated benchmark

4GB
microsoft/Phi-3.5-mini-instructQ4
249.38 tok/sEstimated

Auto-generated benchmark

4GB
deepseek-ai/DeepSeek-R1-Distill-Llama-8BQ4
249.03 tok/sEstimated

Auto-generated benchmark

4GB
mistralai/Mistral-7B-v0.1Q4
247.89 tok/sEstimated

Auto-generated benchmark

4GB
trl-internal-testing/tiny-Qwen2ForCausalLM-2.5Q4
247.60 tok/sEstimated

Auto-generated benchmark

4GB
tencent/HunyuanOCRQ4
247.59 tok/sEstimated

Auto-generated benchmark

1GB
meta-llama/Llama-Guard-3-8BQ4
247.34 tok/sEstimated

Auto-generated benchmark

4GB
GSAI-ML/LLaDA-8B-InstructQ4
247.20 tok/sEstimated

Auto-generated benchmark

4GB
Qwen/Qwen2.5-7BQ4
246.36 tok/sEstimated

Auto-generated benchmark

4GB
Qwen/Qwen2.5-3B-InstructQ4
245.70 tok/sEstimated

Auto-generated benchmark

2GB
Qwen/Qwen3-4BQ4
245.62 tok/sEstimated

Auto-generated benchmark

2GB
trl-internal-testing/tiny-LlamaForCausalLM-3.2Q4
245.56 tok/sEstimated

Auto-generated benchmark

4GB
Qwen/Qwen3-8BQ4
245.44 tok/sEstimated

Auto-generated benchmark

4GB
Qwen/Qwen3-Reranker-0.6BQ4
245.36 tok/sEstimated

Auto-generated benchmark

3GB
Qwen/Qwen3-Embedding-4BQ4
245.00 tok/sEstimated

Auto-generated benchmark

2GB
HuggingFaceTB/SmolLM-135MQ4
244.51 tok/sEstimated

Auto-generated benchmark

4GB
meta-llama/Meta-Llama-3-8BQ4
244.31 tok/sEstimated

Auto-generated benchmark

4GB
black-forest-labs/FLUX.2-devQ4
244.26 tok/sEstimated

Auto-generated benchmark

4GB
microsoft/DialoGPT-smallQ4
243.04 tok/sEstimated

Auto-generated benchmark

4GB
allenai/Olmo-3-7B-ThinkQ4
242.76 tok/sEstimated

Auto-generated benchmark

4GB
deepseek-ai/DeepSeek-OCR
Q4
2GB
305.06 tok/sEstimated
Auto-generated benchmark
google/gemma-2b
Q4
1GB
299.56 tok/sEstimated
Auto-generated benchmark
google-t5/t5-3b
Q4
2GB
292.06 tok/sEstimated
Auto-generated benchmark
unsloth/Llama-3.2-1B-Instruct
Q4
1GB
292.02 tok/sEstimated
Auto-generated benchmark
meta-llama/Llama-3.2-1B
Q4
1GB
291.66 tok/sEstimated
Auto-generated benchmark
nari-labs/Dia2-2B
Q4
2GB
291.41 tok/sEstimated
Auto-generated benchmark
unsloth/Llama-3.2-3B-Instruct
Q4
2GB
288.71 tok/sEstimated
Auto-generated benchmark
meta-llama/Llama-3.2-3B-Instruct
Q4
2GB
288.46 tok/sEstimated
Auto-generated benchmark
ibm-research/PowerMoE-3b
Q4
2GB
286.95 tok/sEstimated
Auto-generated benchmark
bigcode/starcoder2-3b
Q4
2GB
286.26 tok/sEstimated
Auto-generated benchmark
google/gemma-3-1b-it
Q4
1GB
286.00 tok/sEstimated
Auto-generated benchmark
facebook/sam3
Q4
1GB
283.67 tok/sEstimated
Auto-generated benchmark
google-bert/bert-base-uncased
Q4
1GB
279.10 tok/sEstimated
Auto-generated benchmark
inference-net/Schematron-3B
Q4
2GB
278.91 tok/sEstimated
Auto-generated benchmark
google/gemma-2-2b-it
Q4
1GB
278.86 tok/sEstimated
Auto-generated benchmark
unsloth/gemma-3-1b-it
Q4
1GB
272.47 tok/sEstimated
Auto-generated benchmark
meta-llama/Llama-3.2-3B
Q4
2GB
267.77 tok/sEstimated
Auto-generated benchmark
WeiboAI/VibeThinker-1.5B
Q4
1GB
266.87 tok/sEstimated
Auto-generated benchmark
allenai/OLMo-2-0425-1B
Q4
1GB
265.74 tok/sEstimated
Auto-generated benchmark
context-labs/meta-llama-Llama-3.2-3B-Instruct-FP16
Q4
2GB
263.70 tok/sEstimated
Auto-generated benchmark
meta-llama/Llama-3.2-1B-Instruct
Q4
1GB
261.73 tok/sEstimated
Auto-generated benchmark
Qwen/Qwen2.5-3B
Q4
2GB
261.26 tok/sEstimated
Auto-generated benchmark
meta-llama/Llama-Guard-3-1B
Q4
1GB
258.34 tok/sEstimated
Auto-generated benchmark
deepseek-ai/deepseek-coder-1.3b-instruct
Q4
2GB
257.39 tok/sEstimated
Auto-generated benchmark
LiquidAI/LFM2-1.2B
Q4
1GB
255.31 tok/sEstimated
Auto-generated benchmark
apple/OpenELM-1_1B-Instruct
Q4
1GB
253.25 tok/sEstimated
Auto-generated benchmark
google/embeddinggemma-300m
Q4
1GB
253.20 tok/sEstimated
Auto-generated benchmark
TinyLlama/TinyLlama-1.1B-Chat-v1.0
Q4
1GB
253.12 tok/sEstimated
Auto-generated benchmark
ibm-granite/granite-3.3-2b-instruct
Q4
1GB
252.58 tok/sEstimated
Auto-generated benchmark
IlyaGusev/saiga_llama3_8b
Q4
4GB
249.78 tok/sEstimated
Auto-generated benchmark
deepseek-ai/DeepSeek-R1-0528
Q4
4GB
249.76 tok/sEstimated
Auto-generated benchmark
microsoft/Phi-3.5-mini-instruct
Q4
4GB
249.38 tok/sEstimated
Auto-generated benchmark
deepseek-ai/DeepSeek-R1-Distill-Llama-8B
Q4
4GB
249.03 tok/sEstimated
Auto-generated benchmark
mistralai/Mistral-7B-v0.1
Q4
4GB
247.89 tok/sEstimated
Auto-generated benchmark
trl-internal-testing/tiny-Qwen2ForCausalLM-2.5
Q4
4GB
247.60 tok/sEstimated
Auto-generated benchmark
tencent/HunyuanOCR
Q4
1GB
247.59 tok/sEstimated
Auto-generated benchmark
meta-llama/Llama-Guard-3-8B
Q4
4GB
247.34 tok/sEstimated
Auto-generated benchmark
GSAI-ML/LLaDA-8B-Instruct
Q4
4GB
247.20 tok/sEstimated
Auto-generated benchmark
Qwen/Qwen2.5-7B
Q4
4GB
246.36 tok/sEstimated
Auto-generated benchmark
Qwen/Qwen2.5-3B-Instruct
Q4
2GB
245.70 tok/sEstimated
Auto-generated benchmark
Qwen/Qwen3-4B
Q4
2GB
245.62 tok/sEstimated
Auto-generated benchmark
trl-internal-testing/tiny-LlamaForCausalLM-3.2
Q4
4GB
245.56 tok/sEstimated
Auto-generated benchmark
Qwen/Qwen3-8B
Q4
4GB
245.44 tok/sEstimated
Auto-generated benchmark
Qwen/Qwen3-Reranker-0.6B
Q4
3GB
245.36 tok/sEstimated
Auto-generated benchmark
Qwen/Qwen3-Embedding-4B
Q4
2GB
245.00 tok/sEstimated
Auto-generated benchmark
HuggingFaceTB/SmolLM-135M
Q4
4GB
244.51 tok/sEstimated
Auto-generated benchmark
meta-llama/Meta-Llama-3-8B
Q4
4GB
244.31 tok/sEstimated
Auto-generated benchmark
black-forest-labs/FLUX.2-dev
Q4
4GB
244.26 tok/sEstimated
Auto-generated benchmark
microsoft/DialoGPT-small
Q4
4GB
243.04 tok/sEstimated
Auto-generated benchmark
allenai/Olmo-3-7B-Think
Q4
4GB
242.76 tok/sEstimated
Auto-generated benchmark

Note: Performance estimates are calculated. Real results may vary. Methodology · Submit real data

Model compatibility

ModelQuantizationVerdictEstimated speedVRAM needed
openai-community/gpt2FP16Fits comfortably
82.14 tok/sEstimated
15GB (have 40GB)
Qwen/Qwen2.5-7B-InstructQ4Fits comfortably
212.12 tok/sEstimated
4GB (have 40GB)
Qwen/Qwen2.5-7B-InstructQ8Fits comfortably
157.17 tok/sEstimated
7GB (have 40GB)
Qwen/Qwen2.5-7B-InstructFP16Fits comfortably
93.89 tok/sEstimated
15GB (have 40GB)
Qwen/Qwen3-0.6BFP16Fits comfortably
86.25 tok/sEstimated
13GB (have 40GB)
meta-llama/Llama-3.1-8B-InstructQ8Fits comfortably
157.58 tok/sEstimated
9GB (have 40GB)
meta-llama/Llama-3.1-8B-InstructFP16Fits comfortably
80.80 tok/sEstimated
17GB (have 40GB)
dphn/dolphin-2.9.1-yi-1.5-34bQ4Fits comfortably
82.01 tok/sEstimated
17GB (have 40GB)
dphn/dolphin-2.9.1-yi-1.5-34bQ8Fits comfortably
58.89 tok/sEstimated
35GB (have 40GB)
openai/gpt-oss-20bFP16Not supported
46.10 tok/sEstimated
41GB (have 40GB)
google/gemma-3-1b-itQ4Fits comfortably
286.00 tok/sEstimated
1GB (have 40GB)
google/gemma-3-1b-itQ8Fits comfortably
175.67 tok/sEstimated
1GB (have 40GB)
google/gemma-3-1b-itFP16Fits comfortably
96.54 tok/sEstimated
2GB (have 40GB)
Qwen/Qwen3-Embedding-0.6BQ8Fits comfortably
166.01 tok/sEstimated
6GB (have 40GB)
facebook/opt-125mFP16Fits comfortably
83.94 tok/sEstimated
15GB (have 40GB)
TinyLlama/TinyLlama-1.1B-Chat-v1.0Q4Fits comfortably
253.12 tok/sEstimated
1GB (have 40GB)
TinyLlama/TinyLlama-1.1B-Chat-v1.0Q8Fits comfortably
194.52 tok/sEstimated
1GB (have 40GB)
Qwen/Qwen3-4B-Instruct-2507Q8Fits comfortably
171.15 tok/sEstimated
4GB (have 40GB)
Qwen/Qwen3-4B-Instruct-2507FP16Fits comfortably
87.85 tok/sEstimated
9GB (have 40GB)
meta-llama/Llama-3.2-1B-InstructQ4Fits comfortably
261.73 tok/sEstimated
1GB (have 40GB)
meta-llama/Llama-3.2-1B-InstructQ8Fits comfortably
182.83 tok/sEstimated
1GB (have 40GB)
meta-llama/Llama-3.2-1B-InstructFP16Fits comfortably
107.24 tok/sEstimated
2GB (have 40GB)
openai/gpt-oss-120bQ4Not supported
44.76 tok/sEstimated
59GB (have 40GB)
openai/gpt-oss-120bQ8Not supported
34.13 tok/sEstimated
117GB (have 40GB)
context-labs/meta-llama-Llama-3.2-3B-Instruct-FP16Q4Fits comfortably
263.70 tok/sEstimated
2GB (have 40GB)
context-labs/meta-llama-Llama-3.2-3B-Instruct-FP16Q8Fits comfortably
189.76 tok/sEstimated
3GB (have 40GB)
context-labs/meta-llama-Llama-3.2-3B-Instruct-FP16FP16Fits comfortably
99.35 tok/sEstimated
6GB (have 40GB)
mistralai/Mistral-7B-Instruct-v0.2Q4Fits comfortably
230.33 tok/sEstimated
4GB (have 40GB)
mistralai/Mistral-7B-Instruct-v0.2Q8Fits comfortably
143.52 tok/sEstimated
7GB (have 40GB)
mistralai/Mistral-7B-Instruct-v0.2FP16Fits comfortably
86.44 tok/sEstimated
15GB (have 40GB)
Qwen/Qwen3-8BQ8Fits comfortably
160.82 tok/sEstimated
9GB (have 40GB)
Qwen/Qwen3-8BFP16Fits comfortably
78.21 tok/sEstimated
17GB (have 40GB)
inference-net/Schematron-3BQ4Fits comfortably
278.91 tok/sEstimated
2GB (have 40GB)
inference-net/Schematron-3BQ8Fits comfortably
204.68 tok/sEstimated
3GB (have 40GB)
inference-net/Schematron-3BFP16Fits comfortably
109.19 tok/sEstimated
6GB (have 40GB)
deepseek-ai/DeepSeek-R1-Distill-Qwen-32BQ4Fits comfortably
71.99 tok/sEstimated
16GB (have 40GB)
deepseek-ai/DeepSeek-R1-Distill-Qwen-32BQ8Fits comfortably
57.95 tok/sEstimated
33GB (have 40GB)
deepseek-ai/DeepSeek-R1-Distill-Qwen-32BFP16Not supported
30.91 tok/sEstimated
66GB (have 40GB)
Qwen/Qwen2.5-7BFP16Fits comfortably
92.88 tok/sEstimated
15GB (have 40GB)
Qwen/Qwen3-Next-80B-A3B-InstructQ4Fits (tight)
44.99 tok/sEstimated
39GB (have 40GB)
allenai/OLMo-2-0425-1BFP16Fits comfortably
109.38 tok/sEstimated
2GB (have 40GB)
microsoft/Phi-3-mini-4k-instructQ4Fits comfortably
239.34 tok/sEstimated
4GB (have 40GB)
microsoft/Phi-3-mini-4k-instructQ8Fits comfortably
159.70 tok/sEstimated
7GB (have 40GB)
microsoft/Phi-3-mini-4k-instructFP16Fits comfortably
82.87 tok/sEstimated
15GB (have 40GB)
openai-community/gpt2-largeQ8Fits comfortably
144.36 tok/sEstimated
7GB (have 40GB)
openai-community/gpt2-largeFP16Fits comfortably
93.75 tok/sEstimated
15GB (have 40GB)
Qwen/Qwen3-1.7BQ4Fits comfortably
214.33 tok/sEstimated
4GB (have 40GB)
Qwen/Qwen3-1.7BQ8Fits comfortably
148.42 tok/sEstimated
7GB (have 40GB)
Qwen/Qwen3-1.7BFP16Fits comfortably
79.76 tok/sEstimated
15GB (have 40GB)
openai-community/gpt2Q8Fits comfortably
158.89 tok/sEstimated
7GB (have 40GB)
openai-community/gpt2FP16
Fits comfortably15GB required · 40GB available
82.14 tok/sEstimated
Qwen/Qwen2.5-7B-InstructQ4
Fits comfortably4GB required · 40GB available
212.12 tok/sEstimated
Qwen/Qwen2.5-7B-InstructQ8
Fits comfortably7GB required · 40GB available
157.17 tok/sEstimated
Qwen/Qwen2.5-7B-InstructFP16
Fits comfortably15GB required · 40GB available
93.89 tok/sEstimated
Qwen/Qwen3-0.6BFP16
Fits comfortably13GB required · 40GB available
86.25 tok/sEstimated
meta-llama/Llama-3.1-8B-InstructQ8
Fits comfortably9GB required · 40GB available
157.58 tok/sEstimated
meta-llama/Llama-3.1-8B-InstructFP16
Fits comfortably17GB required · 40GB available
80.80 tok/sEstimated
dphn/dolphin-2.9.1-yi-1.5-34bQ4
Fits comfortably17GB required · 40GB available
82.01 tok/sEstimated
dphn/dolphin-2.9.1-yi-1.5-34bQ8
Fits comfortably35GB required · 40GB available
58.89 tok/sEstimated
openai/gpt-oss-20bFP16
Not supported41GB required · 40GB available
46.10 tok/sEstimated
google/gemma-3-1b-itQ4
Fits comfortably1GB required · 40GB available
286.00 tok/sEstimated
google/gemma-3-1b-itQ8
Fits comfortably1GB required · 40GB available
175.67 tok/sEstimated
google/gemma-3-1b-itFP16
Fits comfortably2GB required · 40GB available
96.54 tok/sEstimated
Qwen/Qwen3-Embedding-0.6BQ8
Fits comfortably6GB required · 40GB available
166.01 tok/sEstimated
facebook/opt-125mFP16
Fits comfortably15GB required · 40GB available
83.94 tok/sEstimated
TinyLlama/TinyLlama-1.1B-Chat-v1.0Q4
Fits comfortably1GB required · 40GB available
253.12 tok/sEstimated
TinyLlama/TinyLlama-1.1B-Chat-v1.0Q8
Fits comfortably1GB required · 40GB available
194.52 tok/sEstimated
Qwen/Qwen3-4B-Instruct-2507Q8
Fits comfortably4GB required · 40GB available
171.15 tok/sEstimated
Qwen/Qwen3-4B-Instruct-2507FP16
Fits comfortably9GB required · 40GB available
87.85 tok/sEstimated
meta-llama/Llama-3.2-1B-InstructQ4
Fits comfortably1GB required · 40GB available
261.73 tok/sEstimated
meta-llama/Llama-3.2-1B-InstructQ8
Fits comfortably1GB required · 40GB available
182.83 tok/sEstimated
meta-llama/Llama-3.2-1B-InstructFP16
Fits comfortably2GB required · 40GB available
107.24 tok/sEstimated
openai/gpt-oss-120bQ4
Not supported59GB required · 40GB available
44.76 tok/sEstimated
openai/gpt-oss-120bQ8
Not supported117GB required · 40GB available
34.13 tok/sEstimated
context-labs/meta-llama-Llama-3.2-3B-Instruct-FP16Q4
Fits comfortably2GB required · 40GB available
263.70 tok/sEstimated
context-labs/meta-llama-Llama-3.2-3B-Instruct-FP16Q8
Fits comfortably3GB required · 40GB available
189.76 tok/sEstimated
context-labs/meta-llama-Llama-3.2-3B-Instruct-FP16FP16
Fits comfortably6GB required · 40GB available
99.35 tok/sEstimated
mistralai/Mistral-7B-Instruct-v0.2Q4
Fits comfortably4GB required · 40GB available
230.33 tok/sEstimated
mistralai/Mistral-7B-Instruct-v0.2Q8
Fits comfortably7GB required · 40GB available
143.52 tok/sEstimated
mistralai/Mistral-7B-Instruct-v0.2FP16
Fits comfortably15GB required · 40GB available
86.44 tok/sEstimated
Qwen/Qwen3-8BQ8
Fits comfortably9GB required · 40GB available
160.82 tok/sEstimated
Qwen/Qwen3-8BFP16
Fits comfortably17GB required · 40GB available
78.21 tok/sEstimated
inference-net/Schematron-3BQ4
Fits comfortably2GB required · 40GB available
278.91 tok/sEstimated
inference-net/Schematron-3BQ8
Fits comfortably3GB required · 40GB available
204.68 tok/sEstimated
inference-net/Schematron-3BFP16
Fits comfortably6GB required · 40GB available
109.19 tok/sEstimated
deepseek-ai/DeepSeek-R1-Distill-Qwen-32BQ4
Fits comfortably16GB required · 40GB available
71.99 tok/sEstimated
deepseek-ai/DeepSeek-R1-Distill-Qwen-32BQ8
Fits comfortably33GB required · 40GB available
57.95 tok/sEstimated
deepseek-ai/DeepSeek-R1-Distill-Qwen-32BFP16
Not supported66GB required · 40GB available
30.91 tok/sEstimated
Qwen/Qwen2.5-7BFP16
Fits comfortably15GB required · 40GB available
92.88 tok/sEstimated
Qwen/Qwen3-Next-80B-A3B-InstructQ4
Fits (tight)39GB required · 40GB available
44.99 tok/sEstimated
allenai/OLMo-2-0425-1BFP16
Fits comfortably2GB required · 40GB available
109.38 tok/sEstimated
microsoft/Phi-3-mini-4k-instructQ4
Fits comfortably4GB required · 40GB available
239.34 tok/sEstimated
microsoft/Phi-3-mini-4k-instructQ8
Fits comfortably7GB required · 40GB available
159.70 tok/sEstimated
microsoft/Phi-3-mini-4k-instructFP16
Fits comfortably15GB required · 40GB available
82.87 tok/sEstimated
openai-community/gpt2-largeQ8
Fits comfortably7GB required · 40GB available
144.36 tok/sEstimated
openai-community/gpt2-largeFP16
Fits comfortably15GB required · 40GB available
93.75 tok/sEstimated
Qwen/Qwen3-1.7BQ4
Fits comfortably4GB required · 40GB available
214.33 tok/sEstimated
Qwen/Qwen3-1.7BQ8
Fits comfortably7GB required · 40GB available
148.42 tok/sEstimated
Qwen/Qwen3-1.7BFP16
Fits comfortably15GB required · 40GB available
79.76 tok/sEstimated
openai-community/gpt2Q8
Fits comfortably7GB required · 40GB available
158.89 tok/sEstimated

Note: Performance estimates are calculated. Real results may vary. Methodology · Submit real data

Alternative GPUs

RTX 5070
12GB

Explore how RTX 5070 stacks up for local inference workloads.

RTX 4060 Ti 16GB
16GB

Explore how RTX 4060 Ti 16GB stacks up for local inference workloads.

RX 6800 XT
16GB

Explore how RX 6800 XT stacks up for local inference workloads.

RTX 4070 Super
12GB

Explore how RTX 4070 Super stacks up for local inference workloads.

RTX 3080
10GB

Explore how RTX 3080 stacks up for local inference workloads.