feat(auth): add Kimi and Xiaomi MiMo providers

.github/workflows/ci.yml

@@ -622,6 +622,8 @@ jobs:
           GEMINI_API_KEY: ''
           GEMINI_MODEL: ''
           IDEALAB_API_KEY: ''
+          KIMI_CODE_API_KEY: ''
+          MIMO_API_KEY: ''
           MINIMAX_API_KEY: ''
           MODELSCOPE_API_KEY: ''
           MOONSHOT_API_KEY: ''

docs/design/kimi-xiaomi-auth-providers.md

@@ -0,0 +1,73 @@
+# Kimi and Xiaomi MiMo authentication presets
+
+## Goal
+
+Expose Kimi and Xiaomi MiMo directly under `/auth` → **Third-party
+Providers**, including their subscription endpoints, so users do not have to
+recreate these providers through **Custom Provider**.
+
+## Provider shape
+
+Kimi is represented by one top-level preset with a second-level access-type
+selector:
+
+- **Coding Plan**: the Kimi membership coding benefit at
+  `https://api.kimi.com/coding/v1`, with the four model IDs documented for
+  third-party coding tools.
+- **API Key (China)**: the Kimi Open Platform endpoint at
+  `https://api.moonshot.cn/v1`.
+- **API Key (International)**: the Kimi Open Platform endpoint at
+  `https://api.moonshot.ai/v1`.
+
+The two API Key choices recommend `kimi-k3`, `kimi-k2.7-code`,
+`kimi-k2.7-code-highspeed`, and `kimi-k2.6`. Coding Plan and Kimi Open Platform
+still use different keys, billing, and model IDs. Each endpoint option
+therefore carries its own model list, and the setup flow resolves both the
+model list and environment variable after the access type is selected.
+
+Xiaomi MiMo is represented by one preset because its pay-as-you-go API and
+Token Plan share the same model IDs and OpenAI-compatible request format. Its
+endpoint selector includes pay-as-you-go plus the China, Singapore, and Europe
+Token Plan endpoints. The recommended models are `mimo-v2.5-pro` and
+`mimo-v2.5`.
+
+Both presets remain editable so users can add later model IDs without
+waiting for a Qwen Code release. The Third-party Providers list remains sorted
+alphabetically by its displayed labels.
+
+## Model metadata
+
+- Kimi K3 uses a 1,048,576-token context window, always thinks, and accepts
+  image and video input. Kimi K2.7 Code and K2.6 use a 262,144-token context
+  window and accept image and video input; K2.7 Code always thinks.
+- Kimi Code recommends `k3-256k` first for routine coding. `k3` uses a
+  1,048,576-token context window; all other Kimi Code models use 262,144
+  tokens. `k3` and both K2.7 Code IDs accept image and video input;
+  `k3-256k` accepts images only. Thinking stays enabled so selecting a K3 or
+  K2.7 Code ID does not silently route to K2.6.
+- Both MiMo V2.5 models use a 1,048,576-token context window.
+  `mimo-v2.5` accepts image, video, and audio input; the Pro model is kept
+  text-only according to the published capability table.
+
+No Qwen-specific `enable_thinking` field is installed for Kimi or MiMo. Their
+OpenAI-compatible APIs use different thinking controls and enable thinking by
+default.
+
+## Sources
+
+- [Kimi Code overview](https://www.kimi.com/code/docs/en/)
+- [Kimi Code model configuration](https://www.kimi.com/code/docs/en/kimi-code/models.html)
+- [Kimi API international overview](https://platform.kimi.ai/docs/api/overview)
+- [Kimi API China overview](https://platform.kimi.com/docs/api/overview)
+- [Kimi K3 API guide](https://platform.kimi.ai/docs/guide/kimi-k3-quickstart)
+- [Kimi K2.7 Code API guide](https://platform.kimi.ai/docs/guide/kimi-k2-7-code-quickstart)
+- [Xiaomi MiMo first API call](https://mimo.mi.com/docs/en-US/quick-start/summary/first-api-call)
+- [Xiaomi MiMo Token Plan](https://mimo.mi.com/docs/tokenplan/subscription)
+- [Xiaomi MiMo model overview](https://mimo.mi.com/docs/quick-start/summary/model)
+
+## Verification
+
+Unit tests cover preset metadata, endpoint-specific model resolution,
+install-plan output, registry discovery, and alphabetical ordering. A manual
+E2E pass checks the `/auth` provider list and each endpoint selector without
+submitting real credentials.

docs/users/configuration/auth.md

@@ -3,7 +3,7 @@
 Qwen Code's first-run `/auth` menu has three top-level options. Pick the one that matches how you want to run the CLI:
 
 - **Alibaba ModelStudio**: official recommended setup. Opens a sub-menu with **Coding Plan** (for individual developers · weekly quota included), **Token Plan** (for teams and companies · usage-based billing with a dedicated endpoint), or **Standard API Key** (connect with an existing ModelStudio API key).
-- **Third-party Providers**: choose a built-in provider and connect with an API key (DeepSeek, Grok, MiniMax, Z.AI, Kimi, Idealab, ModelScope, OpenRouter, Requesty).
+- **Third-party Providers**: choose a built-in provider and connect with an API key (DeepSeek, Grok, Idealab, Kimi, MiniMax, ModelScope, OpenRouter, Requesty, Xiaomi MiMo, Z.AI).
 - **Custom Provider**: manually connect a local server, proxy, or unsupported provider — supports OpenAI, Anthropic, Gemini, and other compatible endpoints.
 
 > [!note]