An Anthropic model for careful coding, long-document analysis, and visual understanding with optional adaptive thinking.
Use anthropic/claude-opus-4-8 through the sjolt LLM API.
anthropic/claude-opus-4-8Token rates
anthropic/claude-opus-4-8Playground Examples
Review this JavaScript function for correctness: function average(values) { return values.reduce((sum, value) => sum + value, 0) / values.length; }. Define behavior for empty arrays and non-finite values, suggest a minimal fix, and provide three focused test cases.
Reconcile these project notes into a decision brief. Product: launch on October 12, with CSV export required. Engineering: CSV export needs two weeks after schema approval; the schema is not approved. Operations: launch support must be scheduled ten days ahead. List confirmed facts, conflicts, dependencies, and the next three decisions. Do not invent dates or owners.
Explain how to review a dashboard screenshot for clarity and accessibility. Organize the review around hierarchy, chart labels, contrast, and missing context. Separate directly observable evidence from assumptions, and finish with a reusable checklist.
Similar Models
An Anthropic model for coding, long-context knowledge work, and visual understanding with adaptive thinking.
A fast DeepSeek model for text generation and reasoning tasks.
A DeepSeek model for text generation and reasoning tasks that need more capability.
Model overview
Review code, connect long documents, and interpret images through native Anthropic Messages with optional adaptive thinking.
Supply the relevant implementation, expected behavior, and acceptance criteria. Ask for concrete failure cases and focused changes, and state which findings should be reported.
Use up to 1,000,000 context tokens for specifications, source material, and conversation history. Label sources and request explicit evidence for conclusions.
Attach screenshots, diagrams, or other images alongside your question. The response is text. Page artwork illustrates these workflows and is not output from this language model.
Model characteristics
Up to 1,000,000 context tokens and 128,000 output tokens. max_tokens includes both thinking and the final response.
Enable thinking with thinking.type set to adaptive, or turn it off with disabled. Effort levels are low, medium, high, xhigh, max; the default effort is high. The Playground starts with adaptive thinking enabled.
Send model, messages, and max_tokens to /v1/messages. Put system instructions in the top-level system field. Preserve native content blocks when continuing client-side tool conversations.
State the task, relevant evidence, constraints, and output format. Specify the desired response length, and ask the model to distinguish observations from assumptions.
FAQ
Send model anthropic/claude-opus-4-8 to the Anthropic Messages endpoint /v1/messages with messages and max_tokens. Public API requests can use native streaming.
Yes. Set thinking.type to disabled, or omit thinking to retain the upstream default of off. To enable it, use adaptive and select low, medium, high, xhigh, max through output_config.effort; the default effort is high.
Set max_tokens to a positive integer up to 128,000. This limit covers thinking and final output within the 1,000,000-token context window.
Send text and image attachments, then ask follow-up questions using the visible conversation. The example cards are prompt starters with conceptual illustrations, not recorded model responses.