claude — LLM AI Model
Claude answers a prompt with generated text, and earns its place on multi-step reasoning, code, and documents long enough to defeat smaller models. Set the system prompt, temperature and length. Bring your own API key.
What it does
Claude answers on Anthropic's servers: the text on the node's input port, plus the instruction you set, comes back as generated text. No weights are downloaded and no GPU is reserved, and the call is billed to your own Anthropic key.
In a flow it usually sits between other nodes - turning a vision model's raw output into something a person can read, or expanding a short line into a full prompt for an image or video generator. Tiers run from fastest and cheapest to the heaviest reasoning model; pick by how hard the task is, not by how new the name is.
When to use it
- Reading a vision model's output back as a report a person can use
- Expanding a one-line idea into a full prompt for an image or video generator
- Summarising, translating or restructuring text already moving through a flow
- Multi-step reasoning or code, where the cheapest tier visibly gives up
Details
- CategoryLLM
- Year2026
- LicenseAnthropic Commercial Terms
Compliance & Provenance
- ProviderAnthropic · GPAI
- EU AI Act RiskLimited Risk
- Art. 50 TransparencyRequired — AI-generated outputs are marked
Inputs & Outputs
- Text InputInputstring
Input text to process
- Generated TextOutputstring
LLM generated output
Parameters
- modelstringdefault Claude Sonnet 5options: Claude Fable 5.1, Claude Opus 5, Claude Sonnet 5, +5 more
Model name
Which Claude model to call. Fable 5.1 = for demanding reasoning and long-horizon agentic work, slowest and most expensive. Opus 5 = complex agentic coding and enterprise work. Sonnet 5 (default) = the best combination of speed and intelligence, and cheaper than the Sonnet 4.5 it replaces. Haiku 4.5 = fastest and cheapest, best for simple prompts and high-volume calls. The 4.x Opus and Sonnet entries are the previous generation, kept for flows pinned to them. Pick by task difficulty, not by recency.
- promptstringdefault
System prompt / instruction
Optional system prompt that frames the task (role, style, constraints). The user message comes from the text input pin. Leave blank to use the text input alone.
- max_tokensintegerdefault 2048range 1–8192advanced
Max tokens (1-8192)
Maximum tokens for Claude's response. Default 2048. Lower to constrain length / cost; raise up to 8192 for long-form generation. Output stops at natural EOS regardless of this cap.
- temperaturefloatdefault 0.7range 0–1advanced
Temperature (0-1)
Sampling temperature. 0 = deterministic (best for factual / coding tasks). 0.7 (default) = balanced. 1.0 = maximum diversity Claude supports. Lower for precision, higher for creative writing.
Limitations
- Text in, text out. This node has one text input port - images, audio and files cannot reach it.
- Up to 8192 tokens per response, so a long document comes back summarised rather than in full.
- Runs on your own Anthropic API key. Without one configured in the workspace the node cannot run.