NVIDIA L4 vs NVIDIA T4, AI & Machine Learning Comparison
NVIDIA L4 wins 134 of 154 benchmarks, averaging 53.2% faster.
Both cards were measured first-party on our bench, same suite, same test rig.
What the numbers say
The gap is widest in Sana 1.6B, where NVIDIA L4 leads by 1125% (13.84 vs 1.13 images/min); the closest fight is Qwen2.5 1.5B + 0.5B draft (8% apart); VRAM decides part of this one: NVIDIA L4 runs 138 of our 12 AI workloads while the other card runs 111, models that don't fit score zero.
Benchmark results head-to-head
| Benchmark | NVIDIA L4 | NVIDIA T4 | Difference |
|---|---|---|---|
| Qwen3-4B tok/s | 84 | 67.71 | +24% |
| Llama 3.1 8B tok/s | 50.45 | 35.01 | +44% |
| Qwen2.5-Coder 14B tok/s | 27.46 | 19.76 | +39% |
| Qwen3-32B tok/s | 12.42 | 0 | n/a |
| Llama 3.3 70B tok/s | 0 | 0 | n/a |
| Stable Diffusion XL images/min | 5.18 | 2.36 | +119% |
| FLUX.1 dev images/min | 0 | 0 | n/a |
| FLUX.1 Kontext dev images/min | 0 | 0 | n/a |
| Qwen-Image-Edit images/min | 0 | 0 | n/a |
| Wan 2.2 5B (720p) frames/s | 0.15 | 0 | n/a |
| DeepSeek-R1 Distill Llama 8B tok/s | 50.39 | 36.36 | +39% |
| DeepSeek-R1 Distill 1.5B tok/s | 191.38 | 148.72 | +29% |
| DeepSeek-R1 Distill 14B tok/s | 27.39 | 19.21 | +43% |
| DeepSeek-R1 Distill 7B tok/s | 53.27 | 37.58 | +42% |
| Gemma 3 12B tok/s | 31.01 | 23.12 | +34% |
| Gemma 3 4B tok/s | 82.03 | 65.12 | +26% |
| Gemma 4 12B tok/s | 31.49 | 24.03 | +31% |
| Llama 3.2 1B tok/s | 259.48 | 207.71 | +25% |
| Llama 3.2 3B tok/s | 107.1 | 85.94 | +25% |
| Mistral 7B v0.3 tok/s | 54.02 | 40.01 | +35% |
| Mistral Small 24B tok/s | 17.33 | 11.33 | +53% |
| Phi-4 14B tok/s | 27.24 | 17.46 | +56% |
| Phi-4 Mini 3.8B tok/s | 88.48 | 64.15 | +38% |
| Qwen2.5-Coder 7B tok/s | 53.26 | 37.54 | +42% |
| Qwen3 0.6B tok/s | 357.79 | 263.47 | +36% |
| Qwen3 1.7B tok/s | 177.5 | 139.92 | +27% |
| Qwen3 14B tok/s | 27.61 | 19.31 | +43% |
| Qwen3 30B A3B tok/s | 96.15 | 0 | n/a |
| Qwen3 8B tok/s | 48.95 | 36.51 | +34% |
| SmolLM3 3B tok/s | 112.07 | 86.08 | +30% |
| Codestral 22B tok/s | 18.14 | 12.59 | +44% |
| DeepSeek-R1 Distill 32B tok/s | 12.28 | 0 | n/a |
| Devstral Small 24B tok/s | 17.34 | 11.4 | +52% |
| Dolphin 2.9.1 Yi 1.5 34B tok/s | 11.92 | 0 | n/a |
| Dolphin Mistral 24B Venice tok/s | 17.34 | 11.36 | +53% |
| Dolphin X1 8B tok/s | 50.12 | 35.66 | +41% |
| Dolphin 3.0 Llama 3.1 8B tok/s | 50.24 | 35.68 | +41% |
| Dolphin 3.0 R1 Mistral 24B tok/s | 17.32 | 11.32 | +53% |
| Gemma 3 27B tok/s | 14.18 | 0 | n/a |
| Qwen2.5-Coder 32B tok/s | 12.29 | 0 | n/a |
| Qwen3-Coder 30B A3B tok/s | 99.14 | 0 | n/a |
| QwQ 32B tok/s | 12.28 | 0 | n/a |
| StarCoder2 15B tok/s | 24.4 | 15.66 | +56% |
| BiRefNet images/min | 305.81 | 183.48 | +67% |
| Depth Anything V2 Large images/min | 770.37 | 402.34 | +91% |
| Depth Anything V2 Small images/min | 916.66 | 612.25 | +50% |
| Dolphin X1 Trinity Nano 6B tok/s | 166.6 | 114.21 | +46% |
| SAM ViT-Base images/min | 315.89 | 180.73 | +75% |
| SAM ViT-Huge images/min | 70.46 | 34.91 | +102% |
| Swin2SR 4x Upscaler images/min | 18.23 | 10.53 | +73% |
| Sana 1.6B images/min | 13.84 | 1.13 | +1125% |
| FLUX.1 Schnell images/min | 0 | 0 | n/a |
| PixArt-Sigma XL images/min | 8.08 | 0 | n/a |
| Qwen2.5-1.5B-Instruct tok/s | 191.39 | 148.14 | +29% |
| AuraFlow v0.3 images/min | 1.02 | 0 | n/a |
| gpt-oss-20b tok/s | 91.78 | 63.63 | +44% |
| Kokoro TTS 82M x realtime | 97.35 | 42.88 | +127% |
| MusicGen Small x realtime | 1.01 | 0.92 | +10% |
| Phi-3.5-mini-instruct tok/s | 90.56 | 72.07 | +26% |
| Qwen2.5-3B-Instruct tok/s | 110.84 | 86.04 | +29% |
| Stable Diffusion 3.5 Large images/min | 0 | 0 | n/a |
| Stable Diffusion 3.5 Medium images/min | 2.96 | 0 | n/a |
| Whisper large-v3 x realtime | 70.11 | 44.36 | +58% |
| Florence-2 Base images/min | 174.65 | 148.61 | +18% |
| Florence-2 Large images/min | 93.77 | 83.43 | +12% |
| Qwen-AgentWorld-35B-A3B tok/s | 70.23 | 0 | n/a |
| Meta-Llama-3.1-8B-Instruct tok/s | 50.36 | 37.8 | +33% |
| Qwen2.5-7B-Instruct tok/s | 53.32 | 39.01 | +37% |
| Ornith-1.0-35B tok/s | 70.31 | 0 | n/a |
| Kwaipilot_KAT-Coder-V2.5-Dev tok/s | 79.01 | 0 | n/a |
| Qwen2.5-0.5B-Instruct tok/s | 399.4 | 295.47 | +35% |
| GLM-4.7-Flash tok/s | 75.03 | 0 | n/a |
| DeepSeek-R1-0528-Qwen3-8B tok/s | 49.07 | 37.7 | +30% |
| gemma-2-2b-it tok/s | 114.17 | 92.13 | +24% |
| gemma-3-1b-it tok/s | 216.27 | 156.88 | +38% |
| gemma-3-270m-it tok/s | 493.31 | 362.8 | +36% |
| Ornith-1.0-9B tok/s | 43.35 | 32.93 | +32% |
| Qwen_Qwen3-0.6B tok/s | 356.93 | 259.91 | +37% |
| Qwen2.5-14B-Instruct tok/s | 27.37 | 19.84 | +38% |
| Qwen2.5 1.5B LoRA train tok/s | 3720.9 | 1434.4 | +159% |
| Qwen2.5 7B LoRA train tok/s | 1063.9 | 0 | n/a |
| SmolLM2 1.7B LoRA train tok/s | 3240.2 | 1561.9 | +107% |
| TinyLlama 1.1B LoRA train tok/s | 4756 | 1520.1 | +213% |
| Qwen2.5 1.5B served serve tok/s | 1827 | 1337.1 | +37% |
| Qwen2.5 7B served serve tok/s | 469.9 | 0 | n/a |
| SmolLM2 1.7B served serve tok/s | 1524.9 | 1222.5 | +25% |
| TinyLlama 1.1B served serve tok/s | 2584.3 | 1897.8 | +36% |
| AI21-Jamba-Reasoning-3B tok/s | 107.59 | 86.91 | +24% |
| Olmo-3.1-32B-Think tok/s | 12.27 | 0 | n/a |
| Dolphin-Mistral-24B-Venice-Edition tok/s | 17.2 | 11.33 | +52% |
| DarkIdol-Llama-3.1-8B-Instruct-1.2-Uncensored tok/s | 48.82 | 35.73 | +37% |
| DeepSeek-Coder-V2-Lite tok/s | 110.11 | 83.49 | +32% |
| DeepSeek-R1 Distill 14B (Q3_K_M) tok/s | 27.87 | 16.86 | +65% |
| DeepSeek-R1-Distill-Qwen-32B-abliterated tok/s | 12.26 | 0 | n/a |
| dolphin-2.9-llama3-8b tok/s | 50.04 | 35.1 | +43% |
| EVA-Qwen2.5-14B-v0.2 tok/s | 27.3 | 19.92 | +37% |
| gemma-2-2b-it-abliterated tok/s | 113 | 91.91 | +23% |
| gemma-2-9b tok/s | 33.8 | 28.62 | +18% |
| Gemma 3 12B (Q3_K_M) tok/s | 32.29 | 19.79 | +63% |
| GLM-4.7-Flash-REAP-23B-A3B tok/s | 70.87 | 53.83 | +32% |
| Josiefied-Qwen3-8B-abliterated-v1 tok/s | 48.72 | 35.62 | +37% |
| Hermes-3-Llama-3.2-3B tok/s | 106.14 | 84.99 | +25% |
| SmolLM3-3B tok/s | 110.77 | 86.86 | +28% |
| L3-8B-Stheno-v3.2 tok/s | 50.07 | 36.28 | +38% |
| LFM2.5-1.2B tok/s | 278.53 | 221.66 | +26% |
| LFM2.5-8B-A1B tok/s | 170.2 | 139.03 | +22% |
| Llama-2-7B tok/s | 56.06 | 42.98 | +30% |
| Llama-3.2-3B-Instruct-uncensored tok/s | 101.16 | 86.29 | +17% |
| Phi-4-mini tok/s | 87.21 | 66 | +32% |
| Mistral-7B-Instruct-v0.1 tok/s | 53.6 | 40.63 | +32% |
| Mistral-7B-Instruct-v0.2 tok/s | 53.31 | 41.31 | +29% |
| Mistral-7B-Instruct-v0.3 tok/s | 53.66 | 39.87 | +35% |
| Mistral-Nemo-Instruct-2407 tok/s | 33.01 | 22.98 | +44% |
| NemoMix-Unleashed-12B tok/s | 33.02 | 23.19 | +42% |
| Hermes-4-14B tok/s | 27.48 | 19.02 | +44% |
| phi-2 tok/s | 114.99 | 89.6 | +28% |
| Phi-4 14B (Q3_K_M) tok/s | 27.86 | 16.04 | +74% |
| Qwen3-1.7B tok/s | 176.08 | 135.21 | +30% |
| Qwen3-14B tok/s | 27.48 | 19.44 | +41% |
| Qwen3-30B-A3B tok/s | 95.02 | 0 | n/a |
| Qwen3-4B-Instruct-2507 tok/s | 83.03 | 66.89 | +24% |
| Qwen3-8B tok/s | 48.34 | 37.37 | +29% |
| Qwen1.5-0.5B tok/s | 431.41 | 328.29 | +31% |
| Qwen2-1.5B tok/s | 189.18 | 147.68 | +28% |
| Uncensored tok/s | 27.29 | 18.94 | +44% |
| Qwen2.5-32B tok/s | 12.24 | 0 | n/a |
| Qwen2.5-Coder-0.5B tok/s | 397.1 | 294.96 | +35% |
| Qwen2.5-Coder-1.5B tok/s | 184.6 | 148.51 | +24% |
| Qwen2.5-Coder-14B-Instruct-abliterated tok/s | 27.29 | 18.86 | +45% |
| Qwen2.5-Coder-3B tok/s | 109.42 | 85.74 | +28% |
| Qwen2.5-Coder-7B-Instruct-abliterated tok/s | 53.02 | 37.31 | +42% |
| Qwen3-4B-Instruct-2507 tok/s | 83.13 | 67.45 | +23% |
| Qwen3-4B-Thinking-2507 tok/s | 83.12 | 67 | +24% |
| SmolLM2-135M tok/s | 652.15 | 416.24 | +57% |
| Cydonia-24B-v4.3 tok/s | 17.3 | 11.21 | +54% |
| Codestral 22B (Q3_K_M) tok/s | 17.96 | 11.3 | +59% |
| Mistral Small 24B (Q3_K_M) tok/s | 16.62 | 10.13 | +64% |
| Qwen2.5-Coder 32B (Q3_K_M) tok/s | 12.05 | 0 | n/a |
| Qwen3 30B A3B (Q3_K_M) tok/s | 94.03 | 61.82 | +52% |
| Qwen2.5 1.5B + 0.5B draft x vs solo | 0.968 | 0.9 | +8% |
| DeepSeek-R1-Distill-Llama-70B tok/s | 0 | 0 | n/a |
| Hermes-4-70B tok/s | 0 | 0 | n/a |
| Qwen3-Coder-Next-abliterated tok/s | 0 | 0 | n/a |
| Laguna-XS-2.1 tok/s | 0 | 0 | n/a |
| Llama-3.3-70B-Instruct-abliterated tok/s | 0 | 0 | n/a |
| Meta-Llama-3.1-70B tok/s | 0 | 0 | n/a |
| Nanbeige4.2-3B tok/s | 0 | 0 | n/a |
| Nemotron-3-Nano-30B-A3B tok/s | 0 | 0 | n/a |
| Qwen3-Coder-Next tok/s | 0 | 0 | n/a |
| Qwen3-Next-80B-A3B-Thinking tok/s | 0 | 0 | n/a |
| Qwen2.5-72B tok/s | 0 | 0 | n/a |
| Qwen3-Coder-Next tok/s | 0 | 0 | n/a |
| Qwen3-Next-80B-A3B-Thinking tok/s | 0 | 0 | n/a |
| Qwen3-Next-80B-A3B tok/s | 0 | 0 | n/a |
Whole-job comparison
How long each card takes to finish a complete pipeline, not just one model. NVIDIA L4 is faster on 3 of 4; NVIDIA T4 on 1.
| Workflow | NVIDIA L4 | NVIDIA T4 | Difference |
|---|---|---|---|
| 50-image depth pass | 17 s | 13 s | NVIDIA T4 1.28x faster |
| 500-image masking run | 7.5 min | 14.4 min | NVIDIA L4 1.93x faster |
| 200-product catalogue cutout | 12.1 min | 20.3 min | NVIDIA L4 1.68x faster |
| Full codebase review | 36.4 min | 50.7 min | NVIDIA L4 1.39x faster |
Specifications compared
| NVIDIA L4 | NVIDIA T4 | |
|---|---|---|
| VRAM | 24GB | 16GB |
| Architecture | Ada Lovelace | Turing |
| Memory bandwidth | 300 GB/s | 320 GB/s |
| Boost clock | 2,040 MHz | 1,590 MHz |
| TDP | 72 W | 70 W |
| Launch MSRP | $2,500 | $2,299 |
| Release | 2023-03-21 | 2018-09-13 |
FAQ
- Which is better for ai & machine learning: NVIDIA L4 or NVIDIA T4?
- NVIDIA L4 performs better for ai & machine learning, winning 134 of 154 benchmarks in our suite with an average 53.2% advantage.
- What are the main hardware differences between NVIDIA L4 and NVIDIA T4?
- NVIDIA L4 has 24GB VRAM and a 72W TDP, while NVIDIA T4 has 16GB VRAM and a 70W TDP.
- Does VRAM matter more than speed between NVIDIA L4 and NVIDIA T4?
- For AI, yes, NVIDIA L4 fits 20 more of our 12 workloads than NVIDIA T4. A model that exceeds VRAM doesn't run slower, it doesn't run at all, so the card that fits the model wins that workload outright.
- Where is the biggest performance difference between NVIDIA L4 and NVIDIA T4?
- Sana 1.6B: NVIDIA L4 leads by roughly 1125% (13.84 vs 1.13 images/min) in our testing.
NVIDIA L4 full review · NVIDIA T4 full review · All AI & Machine Learning rankings