image input
Sending images to a multimodal model as content blocks, which it reads and bills as tokens set by the image size.
In more detail
Anthropic estimates an image's cost at about its width times height divided by 750 tokens, and downscales large images, so resize before sending. Images can also carry injected instructions, so treat their text as untrusted.
Example
tokens ≈ (width × height) / 750Related terms
- multimodal model — A model that accepts or produces more than one kind of data, such as text, images and audio.
- content block — One typed piece of a message, such as text, an image, a tool call, a tool result or a thinking block.
- LLM token — The unit of text a language model reads and writes: a word, part of a word, a punctuation mark or a byte sequence from the tokenizer's vocabulary.
Taught in
- Tokens And The Context Window — AI Engineering: LLM Apps, RAG & Agents
- Prompt Injection — AI Engineering: LLM Apps, RAG & Agents
- Guardrails, Privacy And Governance — AI Engineering: LLM Apps, RAG & Agents
- Images, Audio And Speech — AI Engineering: LLM Apps, RAG & Agents