Back to models
Intelligence Index
High setting
40.9#20 of 83measured models
Measured · Artificial Analysis v4.3.2- API price from
- $0.75 in / $3.75 out / 1M tokGoogle · direct API · promotional
- Maximum input
- 1.05M tok
Performance profile
ReasoningRank #2 of 107
95.3%
GPQA DiamondCodingRank #15 of 94
56.6%
SciCodeCost efficiencyRank #50 of 67
32.9pts / $
Intelligence ÷ eval costCapabilities
Maximum input
Modalities
Integration
Reasoning trade-offs
Source (opens in a new tab)Gemini 3.8 Flash
Cost per task (USD)
Gemini 3.8 FlashHigh
- Intelligence Index
- 40.9
- Cost / task
- $1.24
- Evaluation time
- 7.9m
Pricing & access
API route
GoogleAPI
GoogleAPITier & contextStandard · Promotional
USD per 1M tokens
- Input
- $0.75
- Output
- $3.75
- Cached input
- $0.075
Gemini API tool compatibility (opens in a new tab)
- Code execution
- Computer use (preview)
- File search
- Maps
- Web search
- URL context
Benchmarks
Full leaderboardSource (opens in a new tab)47.8%, rank 11 of 12656.6%, rank 15 of 9419.7%, rank 23 of 8781.3%, rank 22 of 1251,436 Elo, rank 27 of 101
Humanity’s Last Exam
Frontier reasoning- 1Claude Opus 5.5Max61.4%
- 2Claude Fable 5.1Max59.1%
- 3Gemini 4 ArgonHigh57.1%
- 8Muse WonderMax50.1%
- 9MiMo-V2.6-ProReasoning49.4%
- 10Muse Spark 1.3Max48.7%
- 11Gemini 3.8 FlashHigh47.8%
- 12Gemini 3.1 Pro PreviewReasoning47.0%
- 13Kimi-K3Max46.9%
- 14Step 5 PreviewReasoning46.5%
Peers use their best measured result on this benchmark.
Alternatives
- Scores higherLing-3.1-flashReasoning
- Intelligence Index
- 41.1+0.2
- Cost / task
- $1.00−20%
- Higher score, lower costGPT-6.1 SolMax
- Intelligence Index
- 51.8+10.9
- Cost / task
- $0.724−42%
- More from GoogleGemini 3.1 Pro PreviewReasoning
- Intelligence Index
- 29.7−11.2
- Cost / task
- $1.30+4%
Each model at its highest-scoring measured setting · Intelligence Index v4.3.2 and its evaluation cost per task.