Google: gemini-3.8-flash
7 benchmarks
Leaderboard position
#49
Overall rank (of 1121 models)
2606
Elo rating
7
Benchmarks
Top 4.4%
Percentile
In Google (of 72 models)
Above: DeepSeek: deepseek-v4.1-flash-1017 #48
Below: SpaceXAI: grok-4.5 #50
Benchmark highlights
Best benchmark
Terminal-Bench v2.1
89.4
Worst benchmark
Agent Arena - Net Improvement (25.09.2026)
4.11
Nearest competitor from another company
The model from another company with the most similar benchmark scores.
appeared as nearest in 1 of 7 benchmarks
1/7
14.3%
1/7
14.3%
1/7
14.3%
1/7
14.3%
1/7
14.3%
1/7
14.3%
1/7
14.3%
| Model | Votes | % of benchmarks |
|---|---|---|
| Qwen: qwen3.8-2.4t-a95b | 1 | 14.3% |
| Z.ai: glm-5.3-flash | 1 | 14.3% |
| OpenAI: gpt-5.6-sol | 1 | 14.3% |
| inclusionAI: ling-3.1-flash | 1 | 14.3% |
| Z.ai: GLM 5.2 | 1 | 14.3% |
| DeepSeek: deepseek-v4.1-flash-1017 | 1 | 14.3% |
| Thinking Machines Lab: inkling-2-small | 1 | 14.3% |
Benchmarks
Terminal-Bench v2.1
89.4
Artificial Analysis Intelligence Index v4.1.1 (Maximum Reasoning)
59
Artificial Analysis Intelligence Index v4.3
41
Artificial Analysis Intelligence Index v4.3.2
41
Artifical Analysis Intelligence Index v5.0
31
Artificial Analysis Intelligence Index v5.0.1
31
Agent Arena - Net Improvement (25.09.2026)
4.11
Predicted scores
Estimated from 25 similar models and the global Elo rating. These are projections, not measured results.