- Intelligence
- 24.9
- Rank 176 of 744 measured models
- Coding
- 42.2
- Rank 151 of 298 measured models
- Output speed
- 94.6 t/s
- Rank 181 of 403 measured models
- Price/1M
- $1.56
- Rank 441 of 659 measured models
- Context window
- 1M
- Rank 84 of 1199 measured models
- Cost per task
- $0.205
- Rank 72 of 193 measured models
Grok 4.3 (High) scores 24.9 on the Artificial Analysis Intelligence Index, ranking 176th of 744 measured models. The closest models on this index are Inkling (Xhigh) (25.0), GPT-5 Codex (High) (24.9), and Claude Sonnet 4.6 (24.7). At $1.56 per million tokens (3:1 input/output blended price), it is cheaper than Inkling (Xhigh) ($1.76), GPT-5 Codex (High) ($3.44), and Claude Sonnet 4.6 ($6). With an output speed of 94.6 t/s, it is faster than Claude Sonnet 4.6 (45.2 t/s), slower than Inkling (Xhigh) (180 t/s) and GPT-5 Codex (High) (166 t/s).
Additional benchmarks21
Design Arena · 3d40.9%
Design Arena · Agenticgamedev28.8%
Design Arena · Agentichtmlslides31.8%
Design Arena · Agenticslides Html31.7%
Design Arena · Agenticslides Python Pptx32.4%
Design Arena · Agenticslides31.9%
Design Arena · Androidnative22.7%
Design Arena · Asciiart43.0%
Design Arena · Codecategories43.9%
Design Arena · Dataviz45.2%
Design Arena · Fullstack28.1%
Design Arena · Gamedev43.6%
Design Arena · Godotgamedev20.7%
Design Arena · Htmlslides28.4%
Design Arena · Mobileapps33.6%
Design Arena · Pptxslides32.5%
Design Arena · Python Pptxslides30.8%
Design Arena · Svg37.2%
Design Arena · Uicomponent45.0%
Design Arena · Webapps43.2%
Design Arena · Website44.2%
Models at a similar level
The closest models on the Intelligence Index, one configuration per family.
- Inkling (Xhigh)
Thinking Machines · $1.76 · 180 t/s
25.0+0.1$1.76180 t/sCompare - GPT-5 Codex (High)
OpenAI · $3.44 · 166 t/s
24.9$3.44166 t/sCompare - Claude Sonnet 4.6
Anthropic · $6 · 45.2 t/s
24.7−0.2$645.2 t/sCompare - GPT-5.1
OpenAI · $3.44 · 112 t/s
24.7−0.2$3.44112 t/sCompare
