Skip to content

AI-native design editor. Open-source Figma alternative.

feat(ai): support MiniMax M3

Danila Poyarkovd57a920646c2
diff --git a/CHANGELOG.md b/CHANGELOG.md
index df04a8c2b..74b88a937 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -13,6 +13,7 @@
### Added
+- Use MiniMax M3 as the default MiniMax model for AI design workflows. (#431)
- Export selections, pages and documents as editable PowerPoint (`.pptx`) files from the File menu, CLI and SDK: text, rectangles, ellipses and lines stay native editable elements, while gradients, masks, blends, vectors and icons fall back to embedded images.
- Figma-style Assets panel browsing with component thumbnails, grid/list views, page grouping, context actions, and drag-to-canvas insertion.
- Import HTML, CSS, Tailwind, and JSX as editable documents from the app, CLI, and SDK, and export standalone browser-ready HTML with compiled CSS and optional external assets.
diff --git a/packages/core/src/constants.ts b/packages/core/src/constants.ts
index cd7096d62..ba0fb76d8 100644
--- a/packages/core/src/constants.ts
+++ b/packages/core/src/constants.ts
@@ -299,9 +299,10 @@ export const AI_PROVIDERS: AIProviderDef[] = [
name: 'MiniMax',
keyPlaceholder: 'API key',
keyURL: 'https://platform.minimax.io/user-center/basic-information/interface-key',
- defaultModel: 'MiniMax-M2.7',
+ defaultModel: 'MiniMax-M3',
models: [
- { id: 'MiniMax-M2.7', name: 'MiniMax-M2.7', tag: 'Best' },
+ { id: 'MiniMax-M3', name: 'MiniMax-M3', tag: 'Best' },
+ { id: 'MiniMax-M2.7', name: 'MiniMax-M2.7' },
{ id: 'MiniMax-M2.7-highspeed', name: 'MiniMax-M2.7-highspeed', tag: 'Fast' },
{ id: 'MiniMax-M2.5', name: 'MiniMax-M2.5' },
{ id: 'MiniMax-M2.5-highspeed', name: 'MiniMax-M2.5 Highspeed', tag: 'Fast' },
diff --git a/packages/docs/programmable/ai-chat.md b/packages/docs/programmable/ai-chat.md
index 96957d76e..4f561a60e 100644
--- a/packages/docs/programmable/ai-chat.md
+++ b/packages/docs/programmable/ai-chat.md
@@ -18,16 +18,16 @@ You can configure multiple reusable models and separately assign models for desi
### Supported Providers
-| Provider | Models | Setup |
-|----------|--------|-------|
-| **OpenRouter** | Claude, GPT, Gemini, DeepSeek, Qwen, and others | API key from [openrouter.ai](https://openrouter.ai) |
-| **Anthropic** | Claude Sonnet 4.6, Claude Opus 4.6 | API key from [console.anthropic.com](https://console.anthropic.com) |
-| **OpenAI** | GPT-5.3 Codex, GPT-4.1, o3, o4-mini | API key from [platform.openai.com](https://platform.openai.com) |
-| **Google AI** | Gemini 3.1 Pro, Gemini 3 Flash | API key from [aistudio.google.dev](https://aistudio.google.dev) |
-| **Z.ai** | GLM-5.1, GLM-5, GLM-4.7, GLM-4.5 family | API key from [docs.z.ai](https://docs.z.ai/devpack/quick-start) |
-| **MiniMax** | MiniMax M2.7, M2.7-highspeed, M2.5, M2.1 | API key from [platform.minimax.io](https://platform.minimax.io/user-center/basic-information/interface-key) |
-| **OpenAI-compatible** | Any endpoint with OpenAI API format | Custom base URL + key. Supports Completions and Responses API toggle. |
-| **Anthropic-compatible** | Any endpoint with Anthropic API format | Custom base URL + key |
+| Provider | Models | Setup |
+| ------------------------ | ----------------------------------------------- | ----------------------------------------------------------------------------------------------------------- |
+| **OpenRouter** | Claude, GPT, Gemini, DeepSeek, Qwen, and others | API key from [openrouter.ai](https://openrouter.ai) |
+| **Anthropic** | Claude Sonnet 4.6, Claude Opus 4.6 | API key from [console.anthropic.com](https://console.anthropic.com) |
+| **OpenAI** | GPT-5.3 Codex, GPT-4.1, o3, o4-mini | API key from [platform.openai.com](https://platform.openai.com) |
+| **Google AI** | Gemini 3.1 Pro, Gemini 3 Flash | API key from [aistudio.google.dev](https://aistudio.google.dev) |
+| **Z.ai** | GLM-5.1, GLM-5, GLM-4.7, GLM-4.5 family | API key from [docs.z.ai](https://docs.z.ai/devpack/quick-start) |
+| **MiniMax** | MiniMax M3, M2.7, M2.7-highspeed, M2.5, M2.1 | API key from [platform.minimax.io](https://platform.minimax.io/user-center/basic-information/interface-key) |
+| **OpenAI-compatible** | Any endpoint with OpenAI API format | Custom base URL + key. Supports Completions and Responses API toggle. |
+| **Anthropic-compatible** | Any endpoint with Anthropic API format | Custom base URL + key |
No backend, no subscription — your key talks directly to the provider.
diff --git a/tests/engine/app/ai/model.test.ts b/tests/engine/app/ai/model.test.ts
index 937525949..9feb4889e 100644
--- a/tests/engine/app/ai/model.test.ts
+++ b/tests/engine/app/ai/model.test.ts
@@ -31,6 +31,38 @@ describe('resolveLanguageModelID', () => {
})
describe('model provider registry', () => {
+ test('offers MiniMax-M3 as the default MiniMax model', () => {
+ const provider = AI_PROVIDERS.find(({ id }) => id === 'minimax')
+ expect(provider?.defaultModel).toBe('MiniMax-M3')
+ expect(provider?.models[0]).toMatchObject({ id: 'MiniMax-M3' })
+ })
+
+ test('sends MiniMax-M3 through the OpenAI-compatible chat endpoint', async () => {
+ let requestURL = ''
+ let requestBody = ''
+ const fetchSpy: typeof fetch = async (input, init) => {
+ requestURL = String(input)
+ requestBody = String(init?.body)
+ throw new Error('stop')
+ }
+ const config: ModelConfig = {
+ providerID: 'minimax',
+ apiKey: 'test-key',
+ modelID: 'MiniMax-M3',
+ customModelID: '',
+ customBaseURL: '',
+ customAPIType: 'completions'
+ }
+
+ const model = modelProviderAdapter('minimax').create(config, { fetch: fetchSpy })
+ await model
+ .doGenerate({ prompt: [{ role: 'user', content: [{ type: 'text', text: 'hi' }] }] })
+ .catch(() => undefined)
+
+ expect(requestURL).toBe('https://api.minimax.io/v1/chat/completions')
+ expect(requestBody).toContain('"model":"MiniMax-M3"')
+ })
+
test('registers every direct provider without handling ACP agents as models', () => {
for (const provider of AI_PROVIDERS) {
expect(modelProviderAdapter(provider.id).create).toBeFunction()