feat: Add Xiaomi MiMo as a first-class API provider (Zoo-Code-Org#81)

* FEATURE: Add Xiaomi MiMo as a first-class API provider

- New MimoHandler with reasoning_content passthrough for multi-turn tool calling
- Custom message conversion preserving MiMo's interleaved thinking chain
- Strip OpenAI-specific extensions (strict, additionalProperties) from tool schemas
- Models: mimo-v2.5-pro, mimo-v2.5, mimo-v2-flash with official pricing
- UI settings with 4 base URL options (3 Token Plan regions + Pay-as-you-go)
- Full wiring: types, schema, model picker, i18n, provider config

* Update openrouter.ts

* Preserve Mimo reasoning fields & switch AMS endpoint

Update Mimo provider config and UI to use the token-plan-ams endpoint instead of the old FRA host. Allow Mimo assistant messages with string content to include and preserve a reasoning_content field when present. Also avoid attaching mapped reasoning_details when converting messages for Mimo models (skip mapping if modelId matches /mimo/i) to preserve the provider's original shape. Types, API handler, transform logic, and the settings UI were updated to keep Mimo-specific reasoning data intact and align the endpoint selection.

* Add MiMo i18n translations to all 17 locales and unit tests

Also hide the "not sure which model" hint for the MiMo provider
since it's a static model list, not fetched dynamically.

* Address CodeRabbit review feedback

Fix Japanese translation consistency, strengthen base URL tests to
verify actual values, and add ModelPicker test for MiMo hint hiding.

* Expand MiMo test coverage and remove dev/null files from PR

Added tests for tool_call_partial streaming, cache token usage,
API error handling, message conversion pipeline, empty delta chunks,
and tools parameter presence/absence. Removed dev/null/ hook files
that were accidentally included in the branch.

* Use data-testid for automaticFetch hint instead of i18n key matching

* Remove mimo-v2-flash model that no longer supports thinking mode

* Replace mimo-v2-flash with mimo-v2-pro model

* Drop mimo-v2-flash since it doesn't support thinking mode

* Address maintainer feedback: revert unrelated changes, add i18n, fix stream_options

* Update doc links to use puter developer URLs

* Fix i18n: JP spacing, FR translations, add multimodal support to message conversion

* Add docstrings and document mimo-v2-flash exclusion rationale

* Address edelauna review: sanitize tool IDs, use handleProviderError, fix prompt cache, clean up comments

* Fold text into last tool message to preserve reasoning continuity

* Refactor: use shared convertToR1Format, processToolCalls, cleanup unused imports/props

* Remove dead temperature param, add 2-tier pricing, clarify strict stripping

* Remove strict/additionalProperties stripping — proxy no longer rejects it

* Use longContextPricing instead of tiers for cost calculation

* Remove stale convertToolsForOpenAI tests (stripping removed)

Author: capitanfeeder

packages/types/src/provider-settings.ts

@@ -20,6 +20,7 @@ import {
 	xaiModels,
 	internationalZAiModels,
 	minimaxModels,
+	mimoModels,
 } from "./providers/index.js"
 
 /**
@@ -121,6 +122,7 @@ export const providerNames = [
 	"mistral",
 	"moonshot",
 	"minimax",
+	"mimo",
 	"openai-codex",
 	"openai-native",
 	"qwen-code",
@@ -334,6 +336,18 @@ const minimaxSchema = apiModelIdProviderModelSchema.extend({
 	minimaxApiKey: z.string().optional(),
 })
 
+const mimoSchema = apiModelIdProviderModelSchema.extend({
+	mimoBaseUrl: z
+		.union([
+			z.literal("https://api.xiaomimimo.com/v1"),
+			z.literal("https://token-plan-cn.xiaomimimo.com/v1"),
+			z.literal("https://token-plan-sgp.xiaomimimo.com/v1"),
+			z.literal("https://token-plan-ams.xiaomimimo.com/v1"),
+		])
+		.optional(),
+	mimoApiKey: z.string().optional(),
+})
+
 const requestySchema = baseProviderSettingsSchema.extend({
 	requestyBaseUrl: z.string().optional(),
 	requestyApiKey: z.string().optional(),
@@ -417,6 +431,7 @@ export const providerSettingsSchemaDiscriminated = z.discriminatedUnion("apiProv
 	poeSchema.merge(z.object({ apiProvider: z.literal("poe") })),
 	moonshotSchema.merge(z.object({ apiProvider: z.literal("moonshot") })),
 	minimaxSchema.merge(z.object({ apiProvider: z.literal("minimax") })),
+	mimoSchema.merge(z.object({ apiProvider: z.literal("mimo") })),
 	requestySchema.merge(z.object({ apiProvider: z.literal("requesty") })),
 	unboundSchema.merge(z.object({ apiProvider: z.literal("unbound") })),
 	fakeAiSchema.merge(z.object({ apiProvider: z.literal("fake-ai") })),
@@ -451,6 +466,7 @@ export const providerSettingsSchema = z.object({
 	...poeSchema.shape,
 	...moonshotSchema.shape,
 	...minimaxSchema.shape,
+	...mimoSchema.shape,
 	...requestySchema.shape,
 	...unboundSchema.shape,
 	...fakeAiSchema.shape,
@@ -525,6 +541,7 @@ export const modelIdKeysByProvider: Record<TypicalProvider, ModelIdKey> = {
 	mistral: "apiModelId",
 	moonshot: "apiModelId",
 	minimax: "apiModelId",
+	mimo: "apiModelId",
 	deepseek: "apiModelId",
 	poe: "apiModelId",
 	"qwen-code": "apiModelId",
@@ -617,6 +634,11 @@ export const MODELS_BY_PROVIDER: Record<
 		label: "MiniMax",
 		models: Object.keys(minimaxModels),
 	},
+	mimo: {
+		id: "mimo",
+		label: "Xiaomi MiMo",
+		models: Object.keys(mimoModels),
+	},
 	"openai-codex": {
 		id: "openai-codex",
 		label: "OpenAI - ChatGPT Plus/Pro",

packages/types/src/providers/index.ts

@@ -25,6 +25,7 @@ export * from "./xai.js"
 export * from "./vercel-ai-gateway.js"
 export * from "./zai.js"
 export * from "./minimax.js"
+export * from "./mimo.js"
 
 import { anthropicDefaultModelId } from "./anthropic.js"
 import { basetenDefaultModelId } from "./baseten.js"
@@ -49,6 +50,7 @@ import { xaiDefaultModelId } from "./xai.js"
 import { vercelAiGatewayDefaultModelId } from "./vercel-ai-gateway.js"
 import { internationalZAiDefaultModelId, mainlandZAiDefaultModelId } from "./zai.js"
 import { minimaxDefaultModelId } from "./minimax.js"
+import { mimoDefaultModelId } from "./mimo.js"
 
 // Import the ProviderName type from provider-settings to avoid duplication
 import type { ProviderName } from "../provider-settings.js"
@@ -85,6 +87,8 @@ export function getProviderDefaultModelId(
 			return moonshotDefaultModelId
 		case "minimax":
 			return minimaxDefaultModelId
+		case "mimo":
+			return mimoDefaultModelId
 		case "zai":
 			return options?.isChina ? mainlandZAiDefaultModelId : internationalZAiDefaultModelId
 		case "openai-native":

packages/types/src/providers/mimo.ts

@@ -0,0 +1,62 @@
+import type { ModelInfo } from "../model.js"
+
+// https://developer.puter.com/ai/xiaomi/mimo-v2.5-pro/
+// https://developer.puter.com/ai/xiaomi/mimo-v2.5/
+// https://platform.xiaomimimo.com/docs/en-US/quick-start/model-hyperparameters
+//
+// NOTE: mimo-v2-flash is not included here. Its thinking mode defaults to
+// disabled and it doesn't reliably handle reasoning_content passthrough
+// during multi-turn tool calling, which causes 400 errors from the proxy.
+// If flash support is needed later, it should be validated against the API
+// first — the tool-calling + thinking flow is what makes MiMo useful as
+// an agentic provider, and flash just can't do that yet.
+export type MimoModelId = keyof typeof mimoModels
+
+export const mimoDefaultModelId: MimoModelId = "mimo-v2.5-pro"
+
+export const mimoModels = {
+	"mimo-v2.5-pro": {
+		maxTokens: 131_072,
+		contextWindow: 1_048_576,
+		supportsImages: false, // Pro series is text-only
+		supportsPromptCache: false,
+		preserveReasoning: true,
+		inputPrice: 1.0, // $1.00/1M tokens (cache miss, ≤256K)
+		outputPrice: 3.0, // $3.00/1M tokens (≤256K)
+		cacheReadsPrice: 0.2, // $0.20/1M tokens (cache hit, ≤256K)
+		cacheWritesPrice: 0, // Free for limited time
+		// MiMo charges 2x above 256K context
+		longContextPricing: {
+			thresholdTokens: 256_000,
+			inputPriceMultiplier: 2,
+			outputPriceMultiplier: 2,
+			cacheReadsPriceMultiplier: 2,
+		},
+		description:
+			"MiMo V2.5 Pro - Xiaomi's flagship reasoning model with 1M context, deep thinking, tool calling, and structured output.",
+	},
+	"mimo-v2.5": {
+		maxTokens: 131_072,
+		contextWindow: 1_048_576,
+		supportsImages: true, // Full-modal: text, image, audio, video input
+		supportsPromptCache: false,
+		preserveReasoning: true,
+		inputPrice: 0.4, // $0.40/1M tokens (cache miss, ≤256K)
+		outputPrice: 2.0, // $2.00/1M tokens (≤256K)
+		cacheReadsPrice: 0.08, // $0.08/1M tokens (cache hit, ≤256K)
+		cacheWritesPrice: 0, // Free for limited time
+		// MiMo charges 2x above 256K context
+		longContextPricing: {
+			thresholdTokens: 256_000,
+			inputPriceMultiplier: 2,
+			outputPriceMultiplier: 2,
+			cacheReadsPriceMultiplier: 2,
+		},
+		description:
+			"MiMo V2.5 - Full-modal understanding model (text, image, audio, video) with 1M context, deep thinking, tool calling, and structured output.",
+	},
+} as const satisfies Record<string, ModelInfo>
+
+export const mimoDefaultModelInfo: ModelInfo = mimoModels[mimoDefaultModelId]
+
+export const MIMO_DEFAULT_TEMPERATURE = 1.0

src/api/index.ts

@@ -33,6 +33,7 @@ import {
 	FireworksHandler,
 	VercelAiGatewayHandler,
 	MiniMaxHandler,
+	MimoHandler,
 	BasetenHandler,
 } from "./providers"
 import { NativeOllamaHandler } from "./providers/native-ollama"
@@ -167,6 +168,8 @@ export function buildApiHandler(configuration: ProviderSettings): ApiHandler {
 			return new LiteLLMHandler(options)
 		case "sambanova":
 			return new SambaNovaHandler(options)
+		case "mimo":
+			return new MimoHandler(options)
 		case "zai":
 			return new ZAiHandler(options)
 		case "fireworks":