Side-by-side comparison of Whisper Large v3 (OpenAI) and Llama 3.1 405B (Meta) — benchmarks, pricing, and capabilities.
| Whisper Large v3 OpenAI | Llama 3.1 405B Meta | |
|---|---|---|
| Category | Audio & Speech | LLMs |
| Specifications | ||
| Context Window | — | 128K |
| Pricing (per 1M tokens) | ||
| Input Cost | — | Free |
| Output Cost | — | Free |
| Performance | ||
| Overall Score | 92.0 | 88.4 |
| ARC-Challenge | — | 87.6 |
| BigBench Hard | — | 82.3 |
| Chatbot Arena ELO | — | 1126.0 |
| DROP | — | 82.0 |
| GSM8K | — | 91.0 |
| HumanEval | — | 85.0 |
| MATH | — | 70.1 |
| MMLU | — | 76.4 |
| TruthfulQA | — | 76.4 |
| WER (inverted) | 94.5 | — |
| WinoGrande | — | 79.4 |
| Community | ||
| User Rating | ★ 4.7 | ★ 4.4 |
| Reviews | 1560 | 420 |