Generate a tools/function-calling request and expected response shape.
—
Plugsky is OpenAI-compatible — see docs and start on the free plan (2 free models, no card).
The Function Calling Tester lets you check how well models handle tool use: whether they select the right function, fill arguments correctly, respect your JSON schema, and return valid call objects. It is for developers building agents, RAG systems, or integrations where a model must drive real code. Compare providers and models against the same tool definitions to see which are reliable enough for production. Plugsky supports function calling over an OpenAI-compatible endpoint.
Function calling lets a model return a structured request to run a function you defined, complete with arguments, instead of free text. Your application executes the function and can send the result back for a final answer.
Define three to five tools with realistic schemas, then run prompts that require each one, including near-misses that should not trigger a call. Score correct selection, valid arguments, and JSON validity across many runs.
Support varies by model and provider. Plugsky exposes tool calling through its OpenAI-compatible chat completions endpoint, so the same client code works across the supported models in its catalogue. Check the docs for current coverage.
Canonical pricing and plans: plugsky.com/#sec-pricing · Terms · SLA · Docs