Key Specifications
| Vendor | anthropic |
|---|
| Version | 3-sonnet-20240229 |
|---|
| Release Date | 2024-02-29 |
|---|
| Context Window | 200000 tokens |
|---|
| Input Modalities | text, image |
|---|
| Output Modalities | text |
|---|
| License | Proprietary |
|---|
| Documentation | https://docs.anthropic.com/claude/docs |
|---|
Benchmark Performance
| Benchmark | Score | Unit | Evaluated At | Notes | Source |
|---|
| MMLU | 84 | % | 2024-02-29 | 5-shot | view |
| HUMANEVAL | 80.1 | pass@1 | 2024-02-29 | — | view |
| GSM8K | 82.6 | % | 2024-02-29 | 0-shot CoT | view |
| MATH | 46.4 | % | 2024-02-29 | 0-shot CoT | view |
| BBH | 76.2 | % | 2024-02-29 | 3-shot CoT | view |
| GPQA | 41.7 | % | 2024-02-29 | 0-shot | view |
| IFEVAL | 75.5 | % | 2024-02-29 | prompt_strict | view |
| ARC | 94.8 | % | 2024-02-29 | challenge | view |
| MUSR | 63.5 | % | 2024-02-29 | 0-shot | view |
| WINOGRANDE | 80.3 | % | 2024-02-29 | 0-shot | view |
Pricing
| Tier | Price | Currency |
|---|
| Input | $3 / Mtok | USD |
| Output | $15 / Mtok | USD |
| Cache Read | $0 / Mtok | USD |
| Cache Write | $0 / Mtok | USD |
Source:
https://www.anthropic.com/pricing
· as of 2024-02-29
Compliance
- Data Residency: US
- SOC2: ✓
- HIPAA: ✗
- GDPR: ✓
- ISO 27001: ✗
Claude 3 Sonnet (2024-02-29)
Modellübersicht
Anthropic Claude 3 Sonnet 2024-02-29 快照版本, 200K 上下文, 速度与智能平衡型。
Kernspezifikationen
| Anbieter | Version | Veröffentlichungsdatum | Kontextfenster | Eingabemodalitäten | Ausgabemodalitäten | Lizenz |
|---|
| Anthropic | 3-sonnet-20240229 | 2024-02-29 | 200K | text, image | text | Proprietary |
Benchmark-Leistung
| Benchmark | Ergebnis | Einheit | Notizen |
|---|
| MMLU (Massive Multitask Language Understanding) | 84.0 | % | 5-shot |
| HumanEval | 80.1 | pass@1 | — |
| GSM8K (Grade School Math 8K) | 82.6 | % | 0-shot CoT |
| MATH | 46.4 | % | 0-shot CoT |
| BBH (BIG-Bench Hard) | 76.2 | % | 3-shot CoT |
| GPQA | 41.7 | % | 0-shot |
| IFEval | 75.5 | % | prompt_strict |
| ARC | 94.8 | % | challenge |
| MUSR | 63.5 | % | 0-shot |
| WinoGrande | 80.3 | % | 0-shot |
Preise
| Eingabe | Ausgabe | Cache-Lesen | Cache-Schreiben |
|---|
| — | — | — | — |
pro Million Token
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。
Schwächen
Anwendungsfälle
Referenzen