Gemini 4 Argon vs Qwen3.8-Max

Gemini 4 Argon leads on the Visual Intelligence Index by 10.3 points, and Gemini 4 Argon scores higher on 6 of the 7 benchmarks where the two differ.

Model

Compared with

At a glance

Measure	Gemini 4 Argon	Qwen3.8-Max
Visual Intelligence Index	84.1%	73.8%
Rank	#1 of 10	#5 of 10
Cost per taskMean over the benchmarks, at list price.	$0.110	$0.151
Latency per taskEach benchmark's median, averaged.	23.9s	66.0s
Tokens per task	54,614	68,932
Input / 1M tokens	$2.00	$2.00
Output / 1M tokens	$10.00	$6.00
Context window	1.05M	1M
Availability	Proprietary	Proprietary

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

Capability	Gemini 4 Argon	Qwen3.8-Max	Lead
Perception	81.7%	72.0%	+9.7
Spatial Reasoning	85.4%	74.3%	+11.2
Temporal Reasoning	84.5%	73.6%	+10.9
Action / Event Understanding	84.7%	71.5%	+13.2
Causal Reasoning	85.4%	78.2%	+7.2
Information Retrieval	82.7%	72.0%	+10.7

Benchmark	Gemini 4 Argon	Qwen3.8-Max	Lead
Video-MME v2	77.0%	60.2%	+16.8
LVBench	86.3%	58.9%	+27.4
Perception Test	89.7%	83.9%	+5.7
NExT-QA	90.0%	87.6%	+2.4
Q-Bench Video	66.9%	69.0%	−2.1
EgoSchema	86.4%	85.2%	+1.2
UCF101-AD	81.8%	56.5%	+25.4

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

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