docs: add Cheaper Inference setup example to OpenAI-Compatible providers

[aiapienthusiast]

Summary

This PR adds a setup example for Cheaper Inference to docs/CONFIGURATION.md.
The example uses the existing OPENAI_COMPATIBLE_* settings.
The change is docs-only.
It follows the pattern of #868.

Cheaper Inference is an OpenAI-compatible LLM gateway.
It serves the Chat Completions API at https://api.cheaperinference.com/v1.

Cheaper Inference is one of the fastest-growing AI routers.
Each model costs 15–60% less than the list price of its lab.

The example pins three models in OPENAI_COMPATIBLE_MODELS.
This keeps the image and video models of the gateway out of the selector.

How tested

I ran Prettier on the changed file.
I checked the API with a live key: model list, chat, streaming, tool calls and JSON schema output with gpt-5.4-mini.

Details

Changed file:

• docs/CONFIGURATION.md: new "Example: Cheaper Inference" block under "OpenAI-Compatible Providers".

Commands:

npx prettier@3.6.2 --check docs/CONFIGURATION.md
# All matched files use Prettier code style!

Notes:

• The new block is after the tool-calling paragraph, so "(as above)" in that paragraph still refers to the block before it.
• lib/utils/registry.ts removes a trailing /v1 and adds it back, so the base URL with /v1 works.
• lib/models/fetch-models.ts does not call /v1/models when OPENAI_COMPATIBLE_MODELS is set.
• The provider name has a space, so the value is in quotes.

[vercel]

@aiapienthusiast is attempting to deploy a commit to the morphic Team on Vercel.

A member of the Team first needs to authorize it.

[miurla]

Thanks for the PR. Cheaper Inference already works through the existing OPENAI_COMPATIBLE_* settings, so no setup change is needed. We're keeping the docs from adding more vendor-specific setup examples, so I'm closing this one.

Written by an agent (Claude Code, Opus 5.5).