Add MiniMax LLM adapter #2156
There are no files selected for viewing
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,45 @@ | ||
| from typing import Any | ||
|
|
||
| from unstract.sdk1.adapters.base1 import BaseAdapter, MiniMaxLLMParameters | ||
| from unstract.sdk1.adapters.enums import AdapterTypes | ||
|
|
||
| DESCRIPTION = ( | ||
| "Adapter for MiniMax's OpenAI-compatible API. " | ||
| "Supply a model name and your MiniMax API key; the endpoint is preconfigured." | ||
| ) | ||
|
|
||
|
|
||
| class MiniMaxLLMAdapter(MiniMaxLLMParameters, BaseAdapter): | ||
| @staticmethod | ||
| def get_id() -> str: | ||
| return "minimax|4f0e4241-2430-4921-81bf-8b2c6040d8d2" | ||
|
|
||
| @staticmethod | ||
| def get_metadata() -> dict[str, Any]: | ||
| return { | ||
| "name": "MiniMax", | ||
| "version": "1.0.0", | ||
| "adapter": MiniMaxLLMAdapter, | ||
| "description": DESCRIPTION, | ||
| "is_active": True, | ||
| } | ||
|
|
||
| @staticmethod | ||
| def get_name() -> str: | ||
| return "MiniMax" | ||
|
|
||
| @staticmethod | ||
| def get_description() -> str: | ||
| return DESCRIPTION | ||
|
|
||
| @staticmethod | ||
| def get_provider() -> str: | ||
| return "minimax" | ||
|
|
||
| @staticmethod | ||
| def get_icon() -> str: | ||
| return "/icons/adapter-icons/MiniMax.png" | ||
|
Contributor
There was a problem hiding this comment. The shipped
Both 29953 bytes. Rendering it confirms it's the OpenRouter fork-arrow mark. Users would see OpenRouter's branding on the MiniMax card in the adapter picker. Needs the real MiniMax logo. |
||
|
|
||
| @staticmethod | ||
| def get_adapter_type() -> AdapterTypes: | ||
| return AdapterTypes.LLM | ||
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,113 @@ | ||
| { | ||
| "title": "MiniMax", | ||
| "type": "object", | ||
| "required": [ | ||
| "adapter_name", | ||
| "api_key", | ||
| "model" | ||
| ], | ||
| "properties": { | ||
| "adapter_name": { | ||
| "type": "string", | ||
| "title": "Name", | ||
| "default": "", | ||
| "description": "Provide a unique name for this adapter instance. Example: minimax-1" | ||
| }, | ||
| "api_key": { | ||
| "type": "string", | ||
| "title": "API Key", | ||
| "format": "password", | ||
| "description": "Your MiniMax API key from [platform.minimax.io](https://platform.minimax.io)." | ||
| }, | ||
| "model": { | ||
| "type": "string", | ||
| "title": "Model", | ||
| "default": "MiniMax-M3", | ||
| "description": "The MiniMax model name. Example: MiniMax-M3" | ||
| }, | ||
| "api_base": { | ||
| "type": "string", | ||
| "format": "url", | ||
| "title": "API Base", | ||
| "default": "https://api.minimax.io/v1", | ||
| "description": "MiniMax OpenAI-compatible endpoint. Pre-filled with the default; change only if MiniMax moves the base URL or you use a regional endpoint." | ||
| }, | ||
| "max_tokens": { | ||
| "type": "number", | ||
| "minimum": 0, | ||
| "multipleOf": 1, | ||
| "title": "Maximum Output Tokens", | ||
| "default": 4096, | ||
| "description": "Maximum number of output tokens to limit LLM replies. Leave it empty to use the provider default." | ||
| }, | ||
| "max_retries": { | ||
| "type": "number", | ||
| "minimum": 0, | ||
| "multipleOf": 1, | ||
| "title": "Max Retries", | ||
| "default": 5, | ||
| "description": "The maximum number of times to retry a request if it fails." | ||
| }, | ||
| "timeout": { | ||
| "type": "number", | ||
| "minimum": 0, | ||
| "multipleOf": 1, | ||
| "title": "Timeout", | ||
| "default": 900, | ||
| "description": "Timeout in seconds." | ||
| }, | ||
| "enable_reasoning": { | ||
| "type": "boolean", | ||
| "title": "Enable Reasoning", | ||
| "default": false, | ||
| "description": "Toggle on for reasoning models to route reasoning_effort via the OpenAI-compatible request body." | ||
| } | ||
|
Comment on lines
+59
to
+64
Contributor
There was a problem hiding this comment.
The adapter routes reasoning through Prompt To Fix With AIThis is a comment left during a code review.
Path: unstract/sdk1/src/unstract/sdk1/adapters/llm1/static/minimax.json
Line: 59-64
Comment:
**`enable_reasoning` toggle silently has no effect on MiniMax**
The adapter routes reasoning through `OpenAICompatibleLLMParameters.validate()`, which sends `reasoning_effort` in `extra_body`. MiniMax's OpenAI-compatible API does not recognise `reasoning_effort`; it exposes deep thinking via a `thinking: {type: "adaptive"}` field instead ([MiniMax docs](https://platform.minimax.io/docs/guides/text-generation)). A user who turns this toggle on will have their `temperature` stripped from the request (potentially changing generation behaviour) while MiniMax silently ignores the unknown `reasoning_effort` key — no actual reasoning is activated. Is there a plan to map `enable_reasoning` → `thinking: {type: "adaptive"}` in MiniMax's request body, or should the `enable_reasoning` section be removed from `minimax.json` until that mapping is implemented?
How can I resolve this? If you propose a fix, please make it concise.
Contributor
There was a problem hiding this comment. Confirmed — this is a real issue, not just a doc mismatch. The reasoning path in Severity is bounded to the opt-in case: it only fires if a user flips the toggle. It won't auto-enable, since Fix: dropping the |
||
| }, | ||
| "allOf": [ | ||
| { | ||
| "if": { | ||
| "properties": { | ||
| "enable_reasoning": { | ||
| "const": true | ||
| } | ||
| }, | ||
| "required": [ | ||
| "enable_reasoning" | ||
| ] | ||
| }, | ||
| "then": { | ||
| "properties": { | ||
| "reasoning_effort": { | ||
| "type": "string", | ||
| "enum": [ | ||
| "low", | ||
| "medium", | ||
| "high" | ||
| ], | ||
| "default": "medium", | ||
| "title": "Reasoning Effort", | ||
| "description": "Sets the reasoning strength when Enable Reasoning is on." | ||
| } | ||
| }, | ||
| "required": [ | ||
| "reasoning_effort" | ||
| ] | ||
| } | ||
| }, | ||
| { | ||
| "if": { | ||
| "properties": { | ||
| "enable_reasoning": { | ||
| "const": false | ||
| } | ||
| }, | ||
| "required": [ | ||
| "enable_reasoning" | ||
| ] | ||
| }, | ||
| "then": { | ||
| "properties": {} | ||
| } | ||
| } | ||
| ] | ||
| } | ||
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Routing through
custom_openai/silently zeroes out cost tracking for MiniMax._validate_branded_openai_compatibledelegates toOpenAICompatibleLLMParameters.validate(), whosevalidate_model()unconditionally prependscustom_openai/.cost_modelis only populated when the endpoint is OpenAI's own (base1.py:483), so it stays unset here, andaudit.py:98ends up callingcost_per_token(model="custom_openai/MiniMax-M3")— which raises, gets swallowed by the bareexcept, and recordscost_in_dollars = 0.0atdebuglevel.LiteLLM already ships a native MiniMax provider, and it prices M3:
So every MiniMax token would bill as $0 in usage tracking, silently.
This is why
OpenRouterLLMParametersextendsBaseChatCompletionParametersinstead ofOpenAICompatibleLLMParameters— its own docstring says so: "Routed through LiteLLM's nativeopenrouter/provider so per-token costs resolve and reasoning params map without provider-specific workarounds." MiniMax should follow OpenRouter's template, not NVIDIA Build's.Worth noting NVIDIA Build's
custom_openai/routing is correct for NVIDIA — LiteLLM prices zeronvidia_nim/chat models (3 entries, all rerankers, all$0.0), so there's nothing to forfeit there. That reason doesn't carry over to MiniMax.Bonus:
MinimaxChatConfig.get_api_base()already returnshttps://api.minimax.io/v1, so going native lets you drop the hardcoded_MINIMAX_API_BASEconstant entirely.