Anthropic: Claude Fable Latest vs Meta: Llama 3.3 70B Instruct
Anthropic: Claude Fable Latest vs Meta: Llama 3.3 70B Instruct: Anthropic: Claude Fable Latest has the larger context window (1,000,000 tokens); Meta: Llama 3.3 70B Instruct is cheaper on input ($0.10 per 1M tokens); Anthropic: Claude Fable Latest supports more input modalities (text, image, file). Data via OpenRouter /models, as of 2026-07-13.
Source: OpenRouter /models · as of 2026-07-13
Side-by-side comparison
| Metric | Anthropic: Claude Fable Latest | Meta: Llama 3.3 70B Instruct |
|---|---|---|
| Context (tokens) | 1,000,000 | 131,072 |
| Input $/M | $10.00 | $0.10 |
| Output $/M | $50.00 | $0.32 |
| Input modalities | text, image, file | text |
| Output modalities | text | text |
| Model id | ~anthropic/claude-fable-latest | meta-llama/llama-3.3-70b-instruct |
Frequently asked questions
Which has a larger context window, Anthropic: Claude Fable Latest or Meta: Llama 3.3 70B Instruct?
Anthropic: Claude Fable Latest has the larger context window at 1,000,000 tokens. (via OpenRouter /models, as of 2026-07-13.)
Which is cheaper, Anthropic: Claude Fable Latest or Meta: Llama 3.3 70B Instruct?
Meta: Llama 3.3 70B Instruct is cheaper on input at $0.10 per 1M tokens. (via OpenRouter /models, as of 2026-07-13.)
What input types do Anthropic: Claude Fable Latest and Meta: Llama 3.3 70B Instruct support?
Anthropic: Claude Fable Latest accepts text, image, file; Meta: Llama 3.3 70B Instruct accepts text.
Model details
Build an optimized prompt for either model with Prompeteer →