N
Manufacturer
NVIDIA
Nemotron-Modellfamilie von NVIDIA.
Best token generation
2.263,7 tok/s
Nemotron-3.5-Lightning-30B-A3B
Best prefill
44.171,1 tok/s
Nemotron-3-Nano-4B
∅ Token generation
242,9 tok/s
Average of 620 runs
∅ Prefill
3.456,4 tok/s
Average of 620 runs
Best TTFT
155 ms
Nemotron-3.5-Lightning-30B-A3B
Best harness rate
80%
Nemotron-Cascade-2-30B-A3B
🏆
Nemotron-3.5-Lightning-30B-A3B · fastest model at 2.263,7 tok/s token generation
Best prefill of this manufacturer: 44.171,1 tok/s (Nemotron-3-Nano-4B)
Best prefill of this manufacturer: 44.171,1 tok/s (Nemotron-3-Nano-4B)
Top models by token generation
Best measured generation throughput per model (tok/s), published performance benchmarks.
Models
All benchmarks →14 assigned
Llama-3.1-Nemotron-Nano-4B-v1.14.5B6 runs
283,4tok/s Ø
Min 110,4Max 419,4
Llama-3.1-Nemotron-Ultra-253B-v1253B53 runs
4,3tok/s Ø
Min 0,1Max 62,1
Llama-3.3-Nemotron-Super-49B-v1.549B49 runs
75,4tok/s Ø
Min 1,1Max 364,1
Nemotron-3-Nano-30B-A3B30B92 runs
336,9tok/s Ø
Min 0,3Max 1.773,5
Nemotron-3-Nano-4B4B67 runs
538,4tok/s Ø
Min 0,3Max 2.182,7
Nemotron-3-Nano-Omni-30B-A3B-Reasoning30B62 runs
401,9tok/s Ø
Min 31,4Max 1.791,5
Nemotron-3-Super-120B-A12B120B59 runs
81,4tok/s Ø
Min 1,9Max 600,9
Nemotron-3-Super-120B-A12B-Q8120B0 runs
No performance data yet
Nemotron-3.5-Lightning-30B-A3B30B33 runs
280,7tok/s Ø
Min 50,7Max 2.263,7
Nemotron-Cascade-2-30B-A3B30B79 runs
359,4tok/s Ø
Min 0,3Max 1.768,1
Nemotron-H-47B-Reasoning-128K47B63 runs
45,1tok/s Ø
Min 1,1Max 203,5
Nemotron-Labs-3-Puzzle-75B-A9B75B0 runs
No performance data yet
NVIDIA-Nemotron-3.5-Super-VL-120B-A12B120B6 runs
445,0tok/s Ø
Min 20,4Max 1.562,3
OpenReasoning-Nemotron-32B32B53 runs
107,9tok/s Ø
Min 1,1Max 481,0