docs(opengateway): describe the offline models as the catalog floor

Address CodeRabbit review of 67350553 on #2488: auto, mimo-v2.6-pro and
mimo-v2.6-flash are the route's catalog floor, not guaranteed picker
entries. A saved profile's own model list or an `availableModels`
allowlist can still narrow what /model shows.

Co-Authored-By: Claude Opus 5.5 <[email protected]>

README.md

@@ -443,7 +443,7 @@ OpenClaude supports multiple providers, but behavior is not identical across all
 - Smaller local models can struggle with long multi-step tool flows
 - Some providers impose lower output caps than the CLI defaults, and OpenClaude adapts where possible
 - AI/ML API uses the OpenAI-compatible route, defaults to `gpt-4o`, and only surfaces chat-capable models from its public catalog
-- Twigpine OpenGateway is the fresh-install startup default and requires an API key from https://twigpine.com/console/gateway/keys. It uses one OpenAI-compatible base URL (`https://api.twigpine.com/v1`; the older `opengateway.gitlawb.com` address still works). `/model` lists the gateway's active models from a cached copy of its live list (refreshed at startup once a day, or on demand with `r` in `/model`; a failed startup refresh keeps the last cached list), plus `auto`, `mimo-v2.6-pro`, and `mimo-v2.6-flash`, which are always listed. Do not pin the base URL to `/v1/xiaomi-mimo`.
+- Twigpine OpenGateway is the fresh-install startup default and requires an API key from https://twigpine.com/console/gateway/keys. It uses one OpenAI-compatible base URL (`https://api.twigpine.com/v1`; the older `opengateway.gitlawb.com` address still works). `/model` lists the gateway's active models from a cached copy of its live list (refreshed at startup once a day, or on demand with `r` in `/model`; a failed startup refresh keeps the last cached list), on top of a catalog floor of `auto`, `mimo-v2.6-pro`, and `mimo-v2.6-flash`; a saved profile's model list or an `availableModels` allowlist can narrow what the picker shows. Do not pin the base URL to `/v1/xiaomi-mimo`.
 - Z.AI GLM Coding Plan uses `https://api.z.ai/api/coding/paas/v4` with `glm-5.2` by default. On that direct route, GLM-5.3-Flash is selectable as `glm-5.3-flash` and accepts image input. Its OpenClaude effort choices are `low`, `high`, and `xhigh`, with `xhigh` requesting Z.AI `max`; GLM-5.3 and the existing GLM-5.2 query controls remain supported. These capabilities are not claimed for gateways that happen to carry the same model name.
 - Xiaomi MiMo uses `api-key` header auth on the direct OpenAI-compatible route and currently does not support `/usage` reporting in OpenClaude
 - GitHub Copilot serializes sub-agent execution by default to reduce Premium Request consumption — see [Agent Routing and Step Limits](docs/agent-routing.md#github-copilot-sub-agent-optimization) for tuning

docs/advanced-setup.md

@@ -335,10 +335,11 @@ caches that list for a day and fetches it again at startup once the cache has
 expired; press `r` in `/model` to refresh it on demand. If the startup refresh
 fails, the picker keeps showing the last cached list, so a model the gateway has
 retired since then can stay listed until a refresh succeeds. A manual refresh
-discards the cached list first, so if it fails only the three models below
-remain until the next successful refresh. `auto` (smart routing),
-`mimo-v2.6-pro`, and `mimo-v2.6-flash` are always listed, including when the
-list has never been fetched.
+discards the cached list first, so if it fails only the catalog floor below
+remains until the next successful refresh. `auto` (smart routing),
+`mimo-v2.6-pro`, and `mimo-v2.6-flash` form the route's catalog floor, kept even
+when the list has never been fetched. A saved provider profile's own model list
+or an `availableModels` allowlist can still narrow what the picker shows.
 
 ### Xiaomi MiMo