Vision Understanding Recommendation: Claude 3 Sonnet (2024-02-29)
Vision Understanding → Claude 3 Sonnet (2024-02-29)
Anwendungsfallübersicht
Claude 3 Sonnet (2024-02-29) by Anthropic is a recommended option for the Vision Understanding use case.
Empfohlene Modelle
| Rang | Anbieter | Kontextfenster | Score |
|---|
| #10 | Anthropic | 200K | 36/100 |
Leistungsanalyse
Benchmark-Leistung
| Benchmark | Ergebnis |
|---|
| MMLU | 84.0 |
| HUMANEVAL | 80.1 |
| GSM8K | 82.6 |
| MATH | 46.4 |
| BBH | 76.2 |
| GPQA | 41.7 |
Stärken
- MMLU score 84.0, strong knowledge reasoning.
- HumanEval 80.1, excellent code generation.
- 支持文本、图像、音频多模态输入。
- 上下文窗口 200K，支持长文本。
Anforderungen
- 需要 anthropic 的 API 密钥
- 输入长度须在 200K 上下文窗口内