docs: add The Grid custom LLM guide

[CastilloLuis]

What

Adds a docs page for The Grid as a custom-llm provider, following the existing guidance that "Vapi supports using any OpenAI-compatible endpoint as the LLM. This includes services like OpenRouter, AnyScale, Together AI".

Two files: the new page and one nav entry under Custom LLMs.

Why the maxTokens note

Most of the page is the standard credential + assistant JSON. The one thing I thought was worth documenting is a footgun I hit while testing:

Vapi's maxTokens defaults to 250, and on The Grid that budget is shared with reasoning tokens — which are counted and billed but never spoken. On one short prompt:

Instrument	completion tokens	of which reasoning	finish_reason
text-standard	202	114	stop
agent-max	250	29	length (truncated)

So at the default, agent-max stops mid-sentence. The page recommends raising maxTokens and keeping the brevity instruction in the system prompt instead.

Verification

Everything in the page was tested against the live API using the OpenAI Node SDK with baseURL set the way Vapi sets it:

• streaming works (stream: true, 246 chunks)
• tool calling returns well-formed tool_calls
• max_tokens is honored
• GET /v1/models returns a spec-shaped {object: "list", data: [...]} with 17 instruments

I'm from The Grid, so treat the product framing as vendor-authored and edit it however you'd like — happy to cut the table or the tier list if it's more detail than these pages usually carry.

[lightsage-app]

Lightsage docs evals

Waiting for the staging docs URL before running evals.

Lightsage will start the selected PR evals automatically when GitHub reports a successful docs deployment for this PR. This usually happens within 15 minutes.

Commit: e5bf8b7
Status: waiting for staging docs URL