- Intelligence
- 39.0
- Rank 74 of 744 measured models
- Coding
- 71.1
- Rank 52 of 298 measured models
- Output speed
- 85.5 t/s
- Rank 209 of 403 measured models
- Price/1M
- $5.63
- Rank 583 of 659 measured models
- Context window
- 1.1M
- Rank 21 of 1199 measured models
GPT-5.4 (Xhigh) scores 39.0 on the Artificial Analysis Intelligence Index, ranking 74th of 744 measured models. The closest models on this index are DeepSeek V4.1 Flash (39.5), Gemini 3.7 Flash (39.1), and Grok 4.5 (High) (38.8). At $5.63 per million tokens (3:1 input/output blended price), it is more expensive than DeepSeek V4.1 Flash ($0.525), Gemini 3.7 Flash ($1.50), and Grok 4.5 (High) ($3). With an output speed of 85.5 t/s, it is faster than Grok 4.5 (High) (58.2 t/s), slower than DeepSeek V4.1 Flash (174 t/s) and Gemini 3.7 Flash (299 t/s).
Additional benchmarks13
Design Arena · 3d42.3%
Design Arena · Androidnative47.4%
Design Arena · Asciiart55.4%
Design Arena · Codecategories52.5%
Design Arena · Dataviz56.6%
Design Arena · Fullstack40.9%
Design Arena · Gamedev57.6%
Design Arena · Godotgamedev46.9%
Design Arena · Mobileapps46.1%
Design Arena · Svg57.8%
Design Arena · Uicomponent57.4%
Design Arena · Webapps40.3%
Design Arena · Website52.5%
Models at a similar level
The closest models on the Intelligence Index, one configuration per family.
- Muse Spark 1.2 (Xhigh)
Meta · $2 · 322 t/s
39.6+0.6$2322 t/sCompare - DeepSeek V4.1 Flash
DeepSeek · $0.525 · 174 t/s
39.5+0.5$0.525174 t/sCompare - Gemini 3.7 Flash
Google · $1.50 · 299 t/s
39.1+0.1$1.50299 t/sCompare - Grok 4.5 (High)
SpaceXSI · $3 · 58.2 t/s
38.8−0.2$358.2 t/sCompare
