Skip to main content
Work with AI models or compare model evaluations. Distinguish a provider that produces an output from a source that reports benchmarks, and choose by the capability or evidence your task needs.

Choose an agent

OpenAI, Google Gemini, MiniMax, and StepFun cover different model-powered creation or reasoning capabilities. Perplexity focuses on research with cited sources. TypeSafe focuses on structured outputs. Artificial Analysis and LMArena provide model-comparison and evaluation information. A provider name does not define the inputs for every operation; check the reference and describe the output or comparison you actually need.

Prepare your brief

  • The outcome you need: a researched answer, an edited image, or another media deliverable.
  • Your source material, reference assets, and the constraints that must remain unchanged.
  • The output format, style, language, and evidence requirements that define a useful result.

Example requests

Research this question using recent sources. Cite the evidence for each major conclusion and flag uncertainty.
Create an image from this brief using the supplied reference. Preserve the product’s appearance and follow these composition constraints.

Choose how to connect

Use Navigator for the Darwin interface or read the MCP guide for Darwin’s shared MCP connection. Each Original’s API reference above describes its request schema and response codes. Provider-specific MCP and A2A addresses will be documented when those deployments are available. For REST requests, authenticate with a Darwin API key carrying tools:execute and supply an Idempotency-Key. Check the response before proceeding, and reuse the same key only for the same request.
Last modified on November 2, 2026