Gemini 4 Argon vs Nemotron 3 Nano Omni 30B-A3B

Gemini 4 Argon leads on the Visual Intelligence Index by 23.8 points, and Gemini 4 Argon scores higher on 7 of the 7 benchmarks where the two differ.

Model

Compared with

At a glance

Measure	Gemini 4 Argon	Nemotron 3 Nano Omni 30B-A3B
Visual Intelligence Index	84.1%	60.3%
Rank	#1 of 10	#9 of 10
Cost per taskMean over the benchmarks, at list price.	$0.110	—
Latency per taskEach benchmark's median, averaged.	23.9s	14.7s
Tokens per task	54,614	35,341
Input / 1M tokens	$2.00	—
Output / 1M tokens	$10.00	—
Context window	1.05M	262K
Availability	Proprietary	Open weights

Each capability index is a weighted mean of the benchmarks that contribute to the capability.

Capability	Gemini 4 Argon	Nemotron 3 Nano Omni 30B-A3B	Lead
Perception	81.7%	61.0%	+20.8
Spatial Reasoning	85.4%	62.0%	+23.4
Temporal Reasoning	84.5%	60.2%	+24.3
Action / Event Understanding	84.7%	55.6%	+29.1
Causal Reasoning	85.4%	63.6%	+21.9
Information Retrieval	82.7%	57.1%	+25.6

Benchmark	Gemini 4 Argon	Nemotron 3 Nano Omni 30B-A3B	Lead
Video-MME v2	77.0%	35.0%	+42.1
LVBench	86.3%	54.2%	+32.1
Perception Test	89.7%	70.3%	+19.4
NExT-QA	90.0%	81.2%	+8.9
Q-Bench Video	66.9%	64.9%	+2.1
EgoSchema	86.4%	65.8%	+20.6
UCF101-AD	81.8%	31.0%	+50.8

4.0/ 5

Reachable1.0: There is an endpoint you can call without hosting anything.
First-party API — supported
Third-party API — supported
Portable0.0: You can run it yourself, and are not tied to one vendor.
Open weights — not supported
Self-hostable — not supported
Multimodal input1.0: It takes the footage directly, rather than frames you extracted.
Video ingestion — supported
Audio — supported
Programmable1.0: Output you can parse, and tools it can call on its own.
Structured output — supported
Tool calling — supported
Operable1.0: Usable interactively and in bulk, not only one call at a time.
Streaming — supported
Batch API — supported

5.0/ 5

Reachable1.0: There is an endpoint you can call without hosting anything.
First-party API — supported
Third-party API — supported
Portable1.0: You can run it yourself, and are not tied to one vendor.
Open weights — supported
Self-hostable — supported
Multimodal input1.0: It takes the footage directly, rather than frames you extracted.
Video ingestion — supported
Audio — supported
Programmable1.0: Output you can parse, and tools it can call on its own.
Structured output — supported
Tool calling — supported
Operable1.0: Usable interactively and in bulk, not only one call at a time.
Streaming — supported
Batch API — supported