- Intelligence
- 40.9
- Rank 61 of 750 measured models
- Coding
- 76.3
- Rank 19 of 298 measured models
- Output speed
- 143 t/s
- Rank 107 of 411 measured models
- Price/1M
- $1.50
- Rank 436 of 671 measured models
- Context window
- 1M
- Rank 86 of 1216 measured models
- Cost per task
- $1.24
- Rank 148 of 199 measured models
Gemini 3.8 Flash (High) scores 40.9 on the Artificial Analysis Intelligence Index, ranking 61st of 750 measured models. The closest models on this index are Ling 3.1 Flash (41.1), Claude Opus 4.7 (40.7), and Qwen3.8 Max (40.2). At $1.50 per million tokens (3:1 input/output blended price), it is cheaper than Claude Opus 4.7 ($10) and Qwen3.8 Max ($3), more expensive than Ling 3.1 Flash ($0.45). With an output speed of 143 t/s, it is faster than Claude Opus 4.7 (54.2 t/s) and Qwen3.8 Max (37.1 t/s), slower than Ling 3.1 Flash (210 t/s).
Additional benchmarks15
Design Arena · 3d49.9%
Design Arena · Androidnative33.5%
Design Arena · Asciiart50.8%
Design Arena · Codecategories50.3%
Design Arena · Dataviz51.0%
Design Arena · Fullstack48.1%
Design Arena · Gamedev51.0%
Design Arena · Godotgamedev53.0%
Design Arena · Mobileapps54.4%
Design Arena · Python Pptxslides40.1%
Design Arena · Spreadsheet40.0%
Design Arena · Svg54.7%
Design Arena · Uicomponent53.8%
Design Arena · Webapps46.2%
Design Arena · Website50.0%
Models at a similar level
The closest models on the Intelligence Index, one configuration per family.
- Claude Opus 4.8 (batch)
Anthropic · — · —
41.8+0.9——Compare - Ling 3.1 Flash
InclusionAI · $0.45 · 210 t/s
41.1+0.2$0.45210 t/sCompare - Claude Opus 4.7
Anthropic · $10 · 54.2 t/s
40.7−0.2$1054.2 t/sCompare - Qwen3.8 Max
Alibaba · $3 · 37.1 t/s
40.2−0.7$337.1 t/sCompare
