From 3ec0463da9fdbf8e8e1827ed4a88917c17ce35b6 Mon Sep 17 00:00:00 2001
From: Peter Steinberger <steipete@gmail.com>
Date: Sat, 4 Apr 2026 12:23:24 +0100
Subject: [PATCH] docs: refresh minimax thinking refs

---
 docs/concepts/model-providers.md        |  4 ++++
 docs/gateway/configuration-reference.md |  4 ++++
 docs/providers/minimax.md               | 11 +++++++++++
 docs/tools/thinking.md                  |  2 ++
 4 files changed, 21 insertions(+)

diff --git a/docs/concepts/model-providers.md b/docs/concepts/model-providers.md
index 5cbe94566e3..6c3a0cf6d34 100644
--- a/docs/concepts/model-providers.md
+++ b/docs/concepts/model-providers.md
@@ -557,6 +557,10 @@ MiniMax is configured via `models.providers` because it uses custom endpoints:
 
 See [/providers/minimax](/providers/minimax) for setup details, model options, and config snippets.
 
+On MiniMax's Anthropic-compatible streaming path, OpenClaw disables thinking by
+default unless you explicitly set it, and `/fast on` rewrites
+`MiniMax-M2.7` to `MiniMax-M2.7-highspeed`.
+
 ### Ollama
 
 Ollama ships as a bundled provider plugin and uses Ollama's native API:
diff --git a/docs/gateway/configuration-reference.md b/docs/gateway/configuration-reference.md
index ed80888110a..11d96134958 100644
--- a/docs/gateway/configuration-reference.md
+++ b/docs/gateway/configuration-reference.md
@@ -2455,6 +2455,10 @@ Set `MINIMAX_API_KEY`. Shortcuts:
 `openclaw onboard --auth-choice minimax-global-api` or
 `openclaw onboard --auth-choice minimax-cn-api`.
 The model catalog now defaults to M2.7 only.
+On the Anthropic-compatible streaming path, OpenClaw disables MiniMax thinking
+by default unless you explicitly set `thinking` yourself. `/fast on` or
+`params.fastMode: true` rewrites `MiniMax-M2.7` to
+`MiniMax-M2.7-highspeed`.
 
 </Accordion>
 
diff --git a/docs/providers/minimax.md b/docs/providers/minimax.md
index 7f03634059a..d09d046e9b2 100644
--- a/docs/providers/minimax.md
+++ b/docs/providers/minimax.md
@@ -143,6 +143,12 @@ openclaw onboard --auth-choice minimax-cn-api
 }
 ```
 
+On the Anthropic-compatible streaming path, OpenClaw now disables MiniMax
+thinking by default unless you explicitly set `thinking` yourself. MiniMax's
+streaming endpoint emits `reasoning_content` in OpenAI-style delta chunks
+instead of native Anthropic thinking blocks, which can leak internal reasoning
+into visible output if left enabled implicitly.
+
 ### MiniMax M2.7 as fallback (example)
 
 **Best for:** keep your strongest latest-generation model as primary, fail over to MiniMax M2.7.
@@ -196,6 +202,11 @@ Current MiniMax auth choices in the wizard/CLI:
 - Model refs are `minimax/<model>`.
 - Default chat model: `MiniMax-M2.7`
 - Alternate chat model: `MiniMax-M2.7-highspeed`
+- On `api: "anthropic-messages"`, OpenClaw injects
+  `thinking: { type: "disabled" }` unless thinking is already explicitly set in
+  params/config.
+- `/fast on` or `params.fastMode: true` rewrites `MiniMax-M2.7` to
+  `MiniMax-M2.7-highspeed` on the Anthropic-compatible stream path.
 - Onboarding and direct API-key setup write explicit model definitions with
   `input: ["text", "image"]` for both M2.7 variants
 - The bundled provider catalog currently exposes the chat refs as text-only
diff --git a/docs/tools/thinking.md b/docs/tools/thinking.md
index 3b99975ca6c..e9b397ae4a3 100644
--- a/docs/tools/thinking.md
+++ b/docs/tools/thinking.md
@@ -21,6 +21,7 @@ title: "Thinking Levels"
   - `highest`, `max` map to `high`.
 - Provider notes:
   - Anthropic Claude 4.6 models default to `adaptive` when no explicit thinking level is set.
+  - MiniMax (`minimax/*`) on the Anthropic-compatible streaming path defaults to `thinking: { type: "disabled" }` unless you explicitly set thinking in model params or request params. This avoids leaked `reasoning_content` deltas from MiniMax's non-native Anthropic stream format.
   - Z.AI (`zai/*`) only supports binary thinking (`on`/`off`). Any non-`off` level is treated as `on` (mapped to `low`).
   - Moonshot (`moonshot/*`) maps `/think off` to `thinking: { type: "disabled" }` and any non-`off` level to `thinking: { type: "enabled" }`. When thinking is enabled, Moonshot only accepts `tool_choice` `auto|none`; OpenClaw normalizes incompatible values to `auto`.
 
@@ -57,6 +58,7 @@ title: "Thinking Levels"
 - For `openai/*`, fast mode maps to OpenAI priority processing by sending `service_tier=priority` on supported Responses requests.
 - For `openai-codex/*`, fast mode sends the same `service_tier=priority` flag on Codex Responses. OpenClaw keeps one shared `/fast` toggle across both auth paths.
 - For direct public `anthropic/*` requests, including OAuth-authenticated traffic sent to `api.anthropic.com`, fast mode maps to Anthropic service tiers: `/fast on` sets `service_tier=auto`, `/fast off` sets `service_tier=standard_only`.
+- For `minimax/*` on the Anthropic-compatible path, `/fast on` (or `params.fastMode: true`) rewrites `MiniMax-M2.7` to `MiniMax-M2.7-highspeed`.
 - Explicit Anthropic `serviceTier` / `service_tier` model params override the fast-mode default when both are set. OpenClaw still skips Anthropic service-tier injection for non-Anthropic proxy base URLs.
 
 ## Verbose directives (/verbose or /v)