Nemotron 3 Nano Omni 30B-A3B vs Qwen3.8-Max

Qwen3.8-Max leads on the Visual Intelligence Index by 13.5 points, and Nemotron 3 Nano Omni 30B-A3B scores higher on 0 of the 7 benchmarks where the two differ.

Model

Compared with

NVIDIANemotron 3 Nano Omni 30B-A3B60.3%Rank #8 AlibabaQwen3.8-Max73.8%Rank #4

At a glance

Measure	Nemotron 3 Nano Omni 30B-A3B	Qwen3.8-Max
Visual Intelligence Index	60.3%	73.8%
Rank	#8 of 9	#4 of 9
Cost per taskMean over the benchmarks, at list price.	—	$0.151
Latency per taskEach benchmark's median, averaged.	14.7s	66.0s
Tokens per task	35,341	68,932
Input / 1M tokens	—	$2.00
Output / 1M tokens	—	$6.00
Context window	262K	1M
Availability	Open weights	Proprietary

How cost and latency are measured

Capabilities

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

Capability	Nemotron 3 Nano Omni 30B-A3B	Qwen3.8-Max	Lead
Perception	61.0%	72.0%	−11.0
Spatial Reasoning	62.0%	74.3%	−12.3
Temporal Reasoning	60.2%	73.6%	−13.4
Action / Event Understanding	55.6%	71.5%	−16.0
Causal Reasoning	63.6%	78.2%	−14.6
Information Retrieval	57.1%	72.0%	−14.9

Benchmarks

Benchmark	Nemotron 3 Nano Omni 30B-A3B	Qwen3.8-Max	Lead
Video-MME v2	35.0%	60.2%	−25.3
LVBench	54.2%	58.9%	−4.7
Perception Test	70.3%	83.9%	−13.7
NExT-QA	81.2%	87.6%	−6.4
Q-Bench Video	64.9%	69.0%	−4.2
EgoSchema	65.8%	85.2%	−19.4
UCF101-AD	31.0%	56.5%	−25.5

Usability

Nemotron 3 Nano Omni 30B-A3B

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

Qwen3.8-Max

3.5/ 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 input0.5: It takes the footage directly, rather than frames you extracted.
Video ingestion — supported
Audio — not 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