Llama 4 Scout is 75.0× cheaper per million tokens (blended). Claude Opus 5 has the larger context window (1,000,000 vs 327,680 tokens, 3.1×). Llama 4 Scout is open-weights; the other is not.
Claude Opus 5 vs Llama 4 Scout
Live catalog fields. Best-in-row is highlighted.
Claude Opus 5
Anthropic · anthropic__claude-opus-5
Llama 4 Scoutopen
Meta · meta-llama__llama-4-scout
Identity
| Field | anthropic__claude-opus-5 | meta-llama__llama-4-scout |
|---|---|---|
| Provider | Anthropic | Meta |
| Slug | anthropic__claude-opus-5 | meta-llama__llama-4-scout |
| Status | live | live |
| Open weights | No | Yes |
| License | — | — |
| HuggingFace | — | meta-llama/Llama-4-Scout-17B-16E-Instruct |
Quality
| Field | anthropic__claude-opus-5 | meta-llama__llama-4-scout |
|---|---|---|
| Quality index (0–100) | 91.5 | — |
Cost
| Field | anthropic__claude-opus-5 | meta-llama__llama-4-scout |
|---|---|---|
| Input $ / 1M tokens | $5.00 | $0.10 |
| Cached input $ / 1M | $0.50 | — |
| Output $ / 1M tokens | $25.00 | $0.30 |
Context
| Field | anthropic__claude-opus-5 | meta-llama__llama-4-scout |
|---|---|---|
| Max context tokens | 1,000,000 | 327,680 |
| Max output tokens | 128,000 | 16,384 |
Modalities
| Field | anthropic__claude-opus-5 | meta-llama__llama-4-scout |
|---|---|---|
| Input | text, image, file | text, image |
| Output | text | text |