Skip to content
Vision Index

Qwen3.8-Max vs Kimi K3

Qwen3.8-Max leads on the Visual Intelligence Index by 1.9 points, and Qwen3.8-Max scores higher on 6 of the 7 benchmarks where the two differ.

AlibabaQwen3.8-Max73.8%Rank #4MoonshotKimi K371.9%Rank #5

At a glance

MeasureQwen3.8-MaxKimi K3
Visual Intelligence Index73.8%71.9%
Rank#4 of 9#5 of 9
Cost per taskMean over the benchmarks, at list price.$0.151$0.114
Latency per taskEach benchmark's median, averaged.66.0s29.2s
Tokens per task68,93231,619
Input / 1M tokens$2.00$3.00
Output / 1M tokens$6.00$15.00
Context window1M1.05M
AvailabilityProprietaryOpen weights

How cost and latency are measured

Capabilities

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

CapabilityQwen3.8-MaxKimi K3
Perception72.0%71.5%
Spatial Reasoning74.3%73.7%
Temporal Reasoning73.6%72.4%
Action / Event Understanding71.5%66.8%
Causal Reasoning78.2%75.2%
Information Retrieval72.0%70.1%

Benchmarks

BenchmarkQwen3.8-MaxKimi K3
Video-MME v260.2%55.2%
LVBench58.9%66.8%
Perception Test83.9%82.4%
NExT-QA87.6%86.3%
Q-Bench Video69.0%67.5%
EgoSchema85.2%78.6%
UCF101-AD56.5%39.4%

Usability

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

Kimi K3

4.0/ 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 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
Operable0.5

Usable interactively and in bulk, not only one call at a time.

  • Streaming — supported
  • Batch API — not supported