- Intelligence
- 16.8
- Rank 285 of 744 measured models
- Coding
- 15.6
- Rank 244 of 298 measured models
- Output speed
- 148 t/s
- Rank 89 of 403 measured models
- Price/1M
- $0.688
- Rank 322 of 659 measured models
- Context window
- 400K
- Rank 357 of 1199 measured models
- Cost per task
- $0.0535
- Rank 36 of 193 measured models
GPT-5 mini (High) scores 16.8 on the Artificial Analysis Intelligence Index, ranking 285th of 744 measured models. The closest models on this index are Doubao Seed Code (16.9), Gemma 4 26B A4B (16.7), and o4 Mini (batch) (16.7). At $0.688 per million tokens (3:1 input/output blended price), it is cheaper than o4 Mini (batch) ($0.963), more expensive than Gemma 4 26B A4B ($0.175). With an output speed of 148 t/s, it is faster than Gemma 4 26B A4B (65.6 t/s).
Additional benchmarks8
Design Arena · 3d36.9%
Design Arena · Asciiart44.7%
Design Arena · Codecategories43.5%
Design Arena · Dataviz43.7%
Design Arena · Gamedev46.5%
Design Arena · Svg45.8%
Design Arena · Uicomponent41.9%
Design Arena · Website44.3%
Models at a similar level
The closest models on the Intelligence Index, one configuration per family.
- Ling-2.6-1T
InclusionAI · $0.85 · —
17.0+0.2$0.85—Compare - Doubao Seed Code
ByteDance · — · —
16.9+0.1——Compare - Gemma 4 26B A4B
Google · $0.175 · 65.6 t/s
16.7−0.1$0.17565.6 t/sCompare - o4 Mini (batch)
OpenAI · $0.963 · —
16.7−0.1$0.963—Compare
