- Intelligence
- 43.6
- Rank 45 of 744 measured models
- Coding
- 76.2
- Rank 21 of 298 measured models
- Output speed
- 43.3 t/s
- Rank 351 of 403 measured models
- Price/1M
- $6
- Rank 587 of 659 measured models
- Context window
- 1M
- Rank 84 of 1199 measured models
- Cost per task
- $2.00
- Rank 162 of 193 measured models
Kimi K3 (Max) scores 43.6 on the Artificial Analysis Intelligence Index, ranking 45th of 744 measured models. The closest models on this index are Grok 4.6 (44.3), Step 5 Preview (43.7), and Claude Haiku 5.5 (43.4). At $6 per million tokens (3:1 input/output blended price), it is more expensive than Grok 4.6 ($3), Step 5 Preview ($1.43), and Claude Haiku 5.5 ($0.20). With an output speed of 43.3 t/s, it is slower than Grok 4.6 (53.3 t/s), Step 5 Preview (90.8 t/s), and Claude Haiku 5.5 (238 t/s).
Additional benchmarks15
Design Arena · 3d67.3%
Design Arena · Agenticgamedev55.0%
Design Arena · Androidnative55.4%
Design Arena · Codecategories62.4%
Design Arena · Dataviz63.2%
Design Arena · Fullstack61.8%
Design Arena · Gamedev60.9%
Design Arena · Godotgamedev48.2%
Design Arena · Htmlslides59.6%
Design Arena · Mobileapps57.0%
Design Arena · Python Pptxslides59.2%
Design Arena · Svg59.2%
Design Arena · Uicomponent60.3%
Design Arena · Webapps59.0%
Design Arena · Website59.2%
Models at a similar level
The closest models on the Intelligence Index, one configuration per family.
- GLM-5.3
Z.AI · $2.15 · 62.5 t/s
44.8+1.2$2.1562.5 t/sCompare - Grok 4.6
SpaceXSI · $3 · 53.3 t/s
44.3+0.7$353.3 t/sCompare - Step 5 Preview
StepFun · $1.43 · 90.8 t/s
43.7+0.1$1.4390.8 t/sCompare - Claude Haiku 5.5
Anthropic · $0.20 · 238 t/s
43.4−0.2$0.20238 t/sCompare
