NVIDIA NIM

Name: NVIDIA NIM
Rating: 5.0 (661 reviews)

NVIDIA NIM provides optimized AI model inference APIs for LLMs, vision, and embedding models through NVIDIA cloud infrastructure.

Supported Models

Model	Speed	Latency	Tests
nvidia/nemotron-3-super-120b-a12b	141869.96 t/s	29.75s	5
openai/gpt-oss-20b	239.61 t/s	10.88s	5
openai/gpt-oss-120b	176.64 t/s	6.57s	60
qwen/qwen3-next-80b-a3b-instruct	118.96 t/s	0.58s	10
qwen/qwen3-next-80b-a3b-instruct	118.96 t/s	0.58s	10
qwen/qwen3-next-80b-a3b-thinking	114.52 t/s	10.51s	10
qwen/qwen3-next-80b-a3b-thinking	114.52 t/s	10.51s	10
meta/llama-4-maverick-17b-128e-instruct	100.41 t/s	0.21s	15
mistralai/mixtral-8x22b-instruct-v0.1	89.66 t/s	0.22s	5
deepseek-ai/deepseek-r1	87.63 t/s	8.96s	15
minimaxai/minimax-m2.1	86.36 t/s	2.88s	30
marin/marin-8b-instruct	84.25 t/s	0.44s	5
stepfun-ai/step-3.5-flash	81.82 t/s	4.79s	10
microsoft/phi-4-mini-flash-reasoning	74.19 t/s	0.46s	5
moonshotai/kimi-k2.5	70.81 t/s	8.11s	10
minimaxai/minimax-m2.5	65.62 t/s	4.40s	25
nvidia/llama-3.3-nemotron-super-49b-v1.5	57.55 t/s	11.11s	10
google/gemma-3-27b-it	57.40 t/s	0.20s	10
ai21labs/jamba-1.5-large-instruct	55.60 t/s	0.29s	10
stockmark/stockmark-2-100b-instruct	55.31 t/s	0.74s	5

Showing 20 of 48 models.

Time	Model	Speed	Latency
Apr 11, 02:08 AM	moonshotai/kimi-k2.5	69.24 t/s	8.16s
Apr 11, 02:01 AM	z-ai/glm5	28.19 t/s	14.62s
Apr 10, 12:07 PM	qwen/qwen3-coder-480b-a35b-instruct	52.06 t/s	1.09s
Apr 10, 12:06 PM	moonshotai/kimi-k2.5	72.37 t/s	8.06s
Apr 10, 11:58 AM	minimaxai/minimax-m2.5	62.31 t/s	2.44s
Apr 8, 05:11 PM	moonshotai/kimi-k2-thinking	30.41 t/s	20.72s
Apr 8, 04:12 PM	moonshotai/kimi-k2-thinking	25.84 t/s	29.77s
Apr 7, 06:19 AM	minimaxai/minimax-m2.5	57.87 t/s	5.92s
Apr 7, 06:14 AM	qwen/qwen3.5-397b-a17b	27.27 t/s	3.21s
Apr 7, 06:08 AM	moonshotai/kimi-k2-thinking	71.21 t/s	7.92s