From aecbb9a2ee643ff6e49bee46a651f4b440ce67a5 Mon Sep 17 00:00:00 2001 From: Octopus Date: Fri, 24 Apr 2026 05:43:05 +0800 Subject: [PATCH] feat: add MiniMax as a chat completion provider (#5452) MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit * feat: add MiniMax as a chat completion provider Add MiniMax (https://www.minimax.io) as a first-class chat completion provider. MiniMax already has TTS integration in SillyTavern; this extends support to LLM chat completions via their OpenAI-compatible API. Supported models: - MiniMax-M2.5 (default) — 204K context - MiniMax-M2.5-highspeed — same capability, faster inference Key implementation details: - Reuses existing SECRET_KEYS.MINIMAX (shared with TTS) - API endpoint: https://api.minimax.io/v1 - Temperature clamped to (0.0, 1.0] as required by MiniMax API - Returns hardcoded model list since MiniMax doesn't expose /v1/models - Full UI integration: model selector, sampler parameters, streaming Co-Authored-By: octo-patch * feat: upgrade MiniMax default model to M2.7 - Add MiniMax-M2.7 and MiniMax-M2.7-highspeed to model list - Set MiniMax-M2.7 as default model - Keep all previous models as alternatives * feat: independent request function, vision support, temp clamping for MiniMax - Extract sendMinimaxRequest() following Chutes pattern (PR #4844) with function calling and JSON Schema structured output support - Clamp temperature to (0.01, 1.0] on backend; limit frontend UI max to 1.0 - Enable image inlining for MiniMax M2.7 model - Add MiniMax to slash-commands model selector and tokenizer mapping - Add minimax_model to default preset * feat: add VLM-based vision support for MiniMax M2.7 M2.7 does not natively accept image input. When images are detected in messages, pre-process them via the MiniMax VLM endpoint (/v1/coding_plan/vlm) to convert images to text descriptions before sending to the chat completions API. Uses the same API key. * feat: add M2-her model to MiniMax provider M2-her is MiniMax's dialogue/roleplay-optimized model with 64K context and 2048 max completion tokens. Text-only (no vision). * feat: add MiniMax China endpoint (minimaxi.com) support Add endpoint selector (Global/China) for MiniMax, mirroring the SiliconFlow pattern. Users can now choose between api.minimax.io (international) and api.minimaxi.com (China domestic). * fix: merge consecutive same-role messages for MiniMax MiniMax API rejects consecutive messages with the same role with error 'invalid chat setting (2013)'. Merge them before sending. * review: address PR feedback on MiniMax provider Backend (src/endpoints/backends/chat-completions.js): - Drop the entire MiniMax VLM image-preprocessing path; vision is no longer advertised for this provider, so M2.7 messages now go straight to /chat/completions without a separate VLM round-trip. - Drop the json_schema -> response_format mapping (MiniMax does not document structured-output support; relying on it was speculative). - Drop the backend temperature clamp; the same clamp now lives in the frontend so the wire payload matches what the user sees. - Drop the MINIMAX branch in /status that returned a hard-coded model list; the frontend hardcodes the same list and bypasses /status via noValidateSources, so the round-trip was wasted. - Add a streaming Transform + non-streaming helper that move ... blocks from delta.content / message.content to reasoning_content. MiniMax M2.x emit chain-of-thought inline in content; without this transform the raw tags leak into the rendered chat. Includes a state machine that holds back partial marker bytes so a marker split across SSE chunks is still detected. Frontend: - public/scripts/openai.js: add MINIMAX to noValidateSources so the key is accepted without a /models call; remove the dead saveModelList branch; clamp temperature to (0.0, 1.0] in createGenerationParameters. - public/scripts/reasoning.js: add MINIMAX to the non-streaming reasoning_content extraction case (the backend transform now produces this field for MiniMax responses). - public/scripts/slash-commands.js: add MINIMAX to the /api enum and add a MiniMax case to /api-url so users can switch endpoint by command. - public/scripts/custom-request.js: pass minimax_endpoint through the override-payload merge alongside the other per-source endpoint fields. - public/scripts/tokenizers.js: stop returning openai_model (which was always a MiniMax model id and thus an unknown tokenizer); fall back to gpt-3.5-turbo for a coarse but functional estimate. - public/scripts/tool-calling.js: add MINIMAX to supportedSources so function-calling settings are exposed. - public/index.html: drop the "-- Connect to the API --" placeholder option from the model select (the model list is hardcoded and always populated); remove minimax from the vision data-source attributes on the inline-media controls. - public/img/minimax.svg: replace the multicolor brand SVG with a single-color currentColor version that matches the other provider icons in the connect panel. * review: drop backend parsing, defer to frontend Per reviewer feedback: SillyTavern's reasoningHandler / reasoning_auto_parse setting already extracts ... blocks on the client side, so the backend doesn't need to rewrite MiniMax responses. Removes the SSE Transform, the non-streaming helper, and the corresponding case in reasoning.js. * fix: remove isImageInliningSupported declaration for MINIMAX * fix: remove MINIMAX from stream reasoning parsing * fix: add to autoconnect logic * fix: add missing MINIMAX models from docs * fix: freq. and pres. pen aren't supported for MINIMAX * fix: use clamp function for adjusting temperature * fix: pass minimax_endpoint from connection profile to ChatCompletionService * fix: update supported APIs in slash command documentation * fix: replace bespoke merge with standard MERGE_TOOLS processing * fix: add data-i18n attributes for headers --------- Co-authored-by: octo-patch Co-authored-by: octo-patch Co-authored-by: Cohee <18619528+Cohee1207@users.noreply.github.com> --- default/content/presets/openai/Default.json | 2 + public/img/minimax.svg | 1 + public/index.html | 33 +++++++- public/scripts/RossAscends-mods.js | 1 + public/scripts/custom-request.js | 2 +- public/scripts/extensions/shared.js | 1 + public/scripts/openai.js | 52 +++++++++++++ public/scripts/secrets.js | 3 +- public/scripts/slash-commands.js | 32 +++++++- public/scripts/tokenizers.js | 5 ++ public/scripts/tool-calling.js | 1 + src/constants.js | 6 ++ src/endpoints/backends/chat-completions.js | 86 ++++++++++++++++++++- 13 files changed, 217 insertions(+), 8 deletions(-) create mode 100644 public/img/minimax.svg diff --git a/default/content/presets/openai/Default.json b/default/content/presets/openai/Default.json index 24115b9c1..db422d1bf 100644 --- a/default/content/presets/openai/Default.json +++ b/default/content/presets/openai/Default.json @@ -10,6 +10,8 @@ "mistralai_model": "mistral-large-latest", "chutes_model": "deepseek-ai/DeepSeek-V3-0324", "chutes_sort_models": "alphabetically", + "minimax_model": "MiniMax-M2.7", + "minimax_endpoint": "global", "electronhub_model": "gpt-4o-mini", "electronhub_sort_models": "alphabetically", "electronhub_group_models": false, diff --git a/public/img/minimax.svg b/public/img/minimax.svg new file mode 100644 index 000000000..1d32449ab --- /dev/null +++ b/public/img/minimax.svg @@ -0,0 +1 @@ +Minimax \ No newline at end of file diff --git a/public/index.html b/public/index.html index c4f05c214..875a1ef3b 100644 --- a/public/index.html +++ b/public/index.html @@ -697,7 +697,7 @@ -
+
Temperature
@@ -749,7 +749,7 @@
-
+
Top P
@@ -1997,7 +1997,7 @@
-
+
+
+

MiniMax API Key

+
+ +
+
+
+ For privacy reasons, your API key will be hidden after you click 'Connect'. +
+

MiniMax Endpoint

+ +

MiniMax Model

+ +

Electron Hub API Key

diff --git a/public/scripts/RossAscends-mods.js b/public/scripts/RossAscends-mods.js index d056c13fe..186ccad0f 100644 --- a/public/scripts/RossAscends-mods.js +++ b/public/scripts/RossAscends-mods.js @@ -408,6 +408,7 @@ function RA_autoconnect(PrevApi) { || (secret_state[SECRET_KEYS.ZAI] && oai_settings.chat_completion_source == chat_completion_sources.ZAI) || (secret_state[SECRET_KEYS.POLLINATIONS] && oai_settings.chat_completion_source === chat_completion_sources.POLLINATIONS) || (secret_state[SECRET_KEYS.WORKERS_AI] && oai_settings.chat_completion_source == chat_completion_sources.WORKERS_AI) + || (secret_state[SECRET_KEYS.MINIMAX] && oai_settings.chat_completion_source == chat_completion_sources.MINIMAX) || (isValidUrl(oai_settings.custom_url) && oai_settings.chat_completion_source == chat_completion_sources.CUSTOM) || (secret_state[SECRET_KEYS.AZURE_OPENAI] && oai_settings.chat_completion_source == chat_completion_sources.AZURE_OPENAI) ) { diff --git a/public/scripts/custom-request.js b/public/scripts/custom-request.js index e7a2a5062..82fbce6db 100644 --- a/public/scripts/custom-request.js +++ b/public/scripts/custom-request.js @@ -591,7 +591,7 @@ export class ChatCompletionService { } // Ensure api-url is properly applied for all sources that accept it - ['custom_url', 'vertexai_region', 'zai_endpoint', 'siliconflow_endpoint'].forEach(field => { + ['custom_url', 'vertexai_region', 'zai_endpoint', 'siliconflow_endpoint', 'minimax_endpoint'].forEach(field => { // The order is: connection profile => CC preset => CC settings overridePayload[field] = overridePayload[field] || settings[field] || oai_settings[field]; }); diff --git a/public/scripts/extensions/shared.js b/public/scripts/extensions/shared.js index 0e11cca85..81224859e 100644 --- a/public/scripts/extensions/shared.js +++ b/public/scripts/extensions/shared.js @@ -448,6 +448,7 @@ export class ConnectionManagerRequestService { vertexai_region: profile['api-url'], zai_endpoint: profile['api-url'], siliconflow_endpoint: profile['api-url'], + minimax_endpoint: profile['api-url'], reverse_proxy: proxyPreset?.url, proxy_password: proxyPreset?.password, custom_prompt_post_processing: profile['prompt-post-processing'], diff --git a/public/scripts/openai.js b/public/scripts/openai.js index fdba275f3..104c25e8d 100644 --- a/public/scripts/openai.js +++ b/public/scripts/openai.js @@ -47,6 +47,7 @@ import { SECRET_KEYS, secret_state, writeSecret } from './secrets.js'; import { getEventSourceStream } from './sse-stream.js'; import { + clamp, createThumbnail, delay, download, @@ -197,6 +198,7 @@ export const chat_completion_sources = { ZAI: 'zai', SILICONFLOW: 'siliconflow', WORKERS_AI: 'workers_ai', + MINIMAX: 'minimax', }; const character_names_behavior = { @@ -270,6 +272,11 @@ export const SILICONFLOW_ENDPOINT = { CN: 'cn', }; +export const MINIMAX_ENDPOINT = { + GLOBAL: 'global', + CN: 'cn', +}; + const sensitiveFields = [ 'reverse_proxy', 'proxy_password', @@ -319,6 +326,8 @@ export const settingsToUpdate = { chutes_sort_models: ['#chutes_sort_models', 'chutes_sort_models', false, true], siliconflow_model: ['#model_siliconflow_select', 'siliconflow_model', false, true], siliconflow_endpoint: ['#siliconflow_endpoint', 'siliconflow_endpoint', false, true], + minimax_model: ['#model_minimax_select', 'minimax_model', false, true], + minimax_endpoint: ['#minimax_endpoint', 'minimax_endpoint', false, true], electronhub_model: ['#model_electronhub_select', 'electronhub_model', false, true], electronhub_sort_models: ['#electronhub_sort_models', 'electronhub_sort_models', false, true], electronhub_group_models: ['#electronhub_group_models', 'electronhub_group_models', false, true], @@ -431,6 +440,8 @@ const default_settings = { chutes_sort_models: 'alphabetically', siliconflow_model: 'deepseek-ai/DeepSeek-V3', siliconflow_endpoint: SILICONFLOW_ENDPOINT.GLOBAL, + minimax_model: 'MiniMax-M2.7', + minimax_endpoint: MINIMAX_ENDPOINT.GLOBAL, electronhub_model: 'gpt-4o-mini', electronhub_sort_models: 'alphabetically', electronhub_group_models: false, @@ -1712,6 +1723,8 @@ export function getChatCompletionModel(settings = null) { return settings.groq_model; case chat_completion_sources.SILICONFLOW: return settings.siliconflow_model; + case chat_completion_sources.MINIMAX: + return settings.minimax_model; case chat_completion_sources.ELECTRONHUB: return settings.electronhub_model; case chat_completion_sources.CHUTES: @@ -2845,6 +2858,14 @@ export async function createGenerationParameters(settings, model, type, messages generate_data.siliconflow_endpoint = settings.siliconflow_endpoint || SILICONFLOW_ENDPOINT.GLOBAL; } + if (settings.chat_completion_source === chat_completion_sources.MINIMAX) { + generate_data.minimax_endpoint = settings.minimax_endpoint || MINIMAX_ENDPOINT.GLOBAL; + // MiniMax requires temperature in (0.0, 1.0]; zero is rejected. + if (Number.isFinite(generate_data.temperature)) { + generate_data.temperature = clamp(generate_data.temperature, Number.EPSILON, 1.0); + } + } + if (settings.chat_completion_source === chat_completion_sources.WORKERS_AI) { generate_data.workers_ai_account_id = settings.workers_ai_account_id; generate_data.top_k = settings.top_k_openai > 0 ? Math.min(Number(settings.top_k_openai), 50) : undefined; @@ -4255,6 +4276,7 @@ async function getStatusOpen() { chat_completion_sources.VERTEXAI, chat_completion_sources.PERPLEXITY, chat_completion_sources.ZAI, + chat_completion_sources.MINIMAX, ]; if (noValidateSources.includes(oai_settings.chat_completion_source)) { let status = t`Key saved; press \"Test Message\" to verify.`; @@ -4312,6 +4334,10 @@ async function getStatusOpen() { data.siliconflow_endpoint = oai_settings.siliconflow_endpoint; } + if (oai_settings.chat_completion_source === chat_completion_sources.MINIMAX) { + data.minimax_endpoint = oai_settings.minimax_endpoint; + } + if (oai_settings.chat_completion_source === chat_completion_sources.WORKERS_AI) { data.workers_ai_account_id = oai_settings.workers_ai_account_id; } @@ -5266,6 +5292,15 @@ async function onModelChange() { oai_settings.siliconflow_model = value; } + if ($(this).is('#model_minimax_select')) { + if (!value) { + console.debug('Null MiniMax model selected. Ignoring.'); + return; + } + console.log('MiniMax model changed to', value); + oai_settings.minimax_model = value; + } + if ($(this).is('#model_electronhub_select')) { if (!value || !hasModelsLoaded) { console.debug('Null ElectronHub model selected. Ignoring.'); @@ -5707,6 +5742,15 @@ async function onModelChange() { $('#temp_openai').attr('max', oai_max_temp).val(oai_settings.temp_openai).trigger('input'); } + if (oai_settings.chat_completion_source === chat_completion_sources.MINIMAX) { + const maxContext = oai_settings.minimax_model === 'M2-her' ? 65536 : 204800; + $('#openai_max_context').attr('max', maxContext); + oai_settings.openai_max_context = Math.min(Number($('#openai_max_context').attr('max')), oai_settings.openai_max_context); + $('#openai_max_context').val(oai_settings.openai_max_context).trigger('input'); + oai_settings.temp_openai = Math.min(claude_max_temp, oai_settings.temp_openai); + $('#temp_openai').attr('max', claude_max_temp).val(oai_settings.temp_openai).trigger('input'); + } + if (oai_settings.chat_completion_source == chat_completion_sources.ZAI) { const maxContext = getZaiMaxContext(oai_settings.zai_model, oai_settings.max_context_unlocked); $('#openai_max_context').attr('max', maxContext); @@ -5780,6 +5824,7 @@ async function onConnectButtonClick(e) { [chat_completion_sources.CHUTES]: { key: SECRET_KEYS.CHUTES, selector: '#api_key_chutes', proxy: false }, [chat_completion_sources.POLLINATIONS]: { key: SECRET_KEYS.POLLINATIONS, selector: '#api_key_pollinations', proxy: false }, [chat_completion_sources.WORKERS_AI]: { key: SECRET_KEYS.WORKERS_AI, selector: '#api_key_workers_ai', proxy: false }, + [chat_completion_sources.MINIMAX]: { key: SECRET_KEYS.MINIMAX, selector: '#api_key_minimax', proxy: false }, }; // Vertex AI Express version - use API key @@ -5845,6 +5890,8 @@ function toggleChatCompletionForms() { $('#model_chutes_select').trigger('change'); } else if (oai_settings.chat_completion_source == chat_completion_sources.SILICONFLOW) { $('#model_siliconflow_select').trigger('change'); + } else if (oai_settings.chat_completion_source == chat_completion_sources.MINIMAX) { + $('#model_minimax_select').trigger('change'); } else if (oai_settings.chat_completion_source == chat_completion_sources.ELECTRONHUB) { $('#model_electronhub_select').trigger('change'); } else if (oai_settings.chat_completion_source == chat_completion_sources.NANOGPT) { @@ -7028,6 +7075,10 @@ export function initOpenAI() { oai_settings.siliconflow_endpoint = String($(this).val()); saveSettingsDebounced(); }); + $('#minimax_endpoint').on('input', function () { + oai_settings.minimax_endpoint = String($(this).val()); + saveSettingsDebounced(); + }); $('#workers_ai_account_id').on('input', function () { oai_settings.workers_ai_account_id = String($(this).val()); saveSettingsDebounced(); @@ -7048,6 +7099,7 @@ export function initOpenAI() { $('#model_groq_select').on('change', onModelChange); $('#model_chutes_select').on('change', onModelChange); $('#model_siliconflow_select').on('change', onModelChange); + $('#model_minimax_select').on('change', onModelChange); $('#model_electronhub_select').on('change', onModelChange); $('#model_nanogpt_select').on('change', onModelChange); $('#model_deepseek_select').on('change', onModelChange); diff --git a/public/scripts/secrets.js b/public/scripts/secrets.js index 01788393c..57c8fc410 100644 --- a/public/scripts/secrets.js +++ b/public/scripts/secrets.js @@ -130,7 +130,7 @@ const FRIENDLY_NAMES = { [SECRET_KEYS.LINGVA_URL]: 'Lingva Endpoint (e.g. https://lingva.ml/api/v1)', [SECRET_KEYS.ONERING_URL]: 'OneRingTranslator Endpoint (e.g. http://127.0.0.1:4990/translate)', [SECRET_KEYS.DEEPLX_URL]: 'DeepLX Endpoint (e.g. http://127.0.0.1:1188/translate)', - [SECRET_KEYS.MINIMAX]: 'MiniMax TTS', + [SECRET_KEYS.MINIMAX]: 'MiniMax', [SECRET_KEYS.MINIMAX_GROUP_ID]: 'MiniMax Group ID', [SECRET_KEYS.MOONSHOT]: 'Moonshot AI', [SECRET_KEYS.COMETAPI]: 'CometAPI', @@ -184,6 +184,7 @@ const INPUT_MAP = { [SECRET_KEYS.AZURE_OPENAI]: '#api_key_azure_openai', [SECRET_KEYS.ZAI]: '#api_key_zai', [SECRET_KEYS.SILICONFLOW]: '#api_key_siliconflow', + [SECRET_KEYS.MINIMAX]: '#api_key_minimax', [SECRET_KEYS.POLLINATIONS]: '#api_key_pollinations', [SECRET_KEYS.WORKERS_AI]: '#api_key_workers_ai', }; diff --git a/public/scripts/slash-commands.js b/public/scripts/slash-commands.js index 5dd24aeef..47252bf14 100644 --- a/public/scripts/slash-commands.js +++ b/public/scripts/slash-commands.js @@ -70,7 +70,7 @@ import { hideChatMessageRange } from './chats.js'; import { getContext, saveMetadataDebounced } from './extensions.js'; import { getRegexedString, regex_placement } from './extensions/regex/engine.js'; import { findGroupMemberId, groups, is_group_generating, openGroupById, regenerateGroup, resetSelectedGroup, saveGroupChat, selected_group, getGroupMembers } from './group-chats.js'; -import { chat_completion_sources, oai_settings, promptManager, SILICONFLOW_ENDPOINT, ZAI_ENDPOINT } from './openai.js'; +import { chat_completion_sources, MINIMAX_ENDPOINT, oai_settings, promptManager, SILICONFLOW_ENDPOINT, ZAI_ENDPOINT } from './openai.js'; import { user_avatar } from './personas.js'; import { addEphemeralStoppingString, chat_styles, context_presets, flushEphemeralStoppingStrings, playMessageSound, power_user } from './power-user.js'; import { SERVER_INPUTS, textgen_types, textgenerationwebui_settings } from './textgen-settings.js'; @@ -3134,6 +3134,7 @@ export function initDefaultSlashCommands() { new SlashCommandEnumValue('zai', 'Z.AI', enumTypes.getBasedOnIndex(UNIQUE_APIS.findIndex(x => x === 'zai')), 'Z'), new SlashCommandEnumValue('vertexai', 'Google Vertex AI', enumTypes.getBasedOnIndex(UNIQUE_APIS.findIndex(x => x === 'vertexai')), 'V'), new SlashCommandEnumValue('siliconflow', 'SiliconFlow', enumTypes.getBasedOnIndex(UNIQUE_APIS.findIndex(x => x === 'siliconflow')), 'S'), + new SlashCommandEnumValue('minimax', 'MiniMax', enumTypes.getBasedOnIndex(UNIQUE_APIS.findIndex(x => x === 'minimax')), 'M'), new SlashCommandEnumValue('kobold', 'KoboldAI Classic', enumTypes.getBasedOnIndex(UNIQUE_APIS.findIndex(x => x === 'kobold')), 'K'), ...Object.values(textgen_types).filter(api => Object.keys(SERVER_INPUTS).includes(api)).map(api => new SlashCommandEnumValue(api, null, enumTypes.getBasedOnIndex(UNIQUE_APIS.findIndex(x => x === 'textgenerationwebui')), 'T')), ], @@ -3167,7 +3168,7 @@ export function initDefaultSlashCommands() { ${t`If a manual API is provided to set the URL, make sure to set connect=false, as auto-connect only works for the currently selected API, or consider switching to it with /api first.`}
- ${t`This slash command works for most of the Text Completion sources, KoboldAI Classic, and also Custom OpenAI compatible, Z.AI, SiliconFlow, and Google Vertex AI for the Chat Completion sources. If unsure which APIs are supported, check the auto-completion of the optional api argument of this command.`} + ${t`This slash command works for most of the Text Completion sources, KoboldAI Classic, and also Custom OpenAI compatible, Z.AI, SiliconFlow, MiniMax, and Google Vertex AI for the Chat Completion sources. If unsure which APIs are supported, check the auto-completion of the optional api argument of this command.`}
`, })); @@ -6261,6 +6262,7 @@ function getModelOptions(quiet) { { id: 'model_groq_select', api: 'openai', type: chat_completion_sources.GROQ }, { id: 'model_chutes_select', api: 'openai', type: chat_completion_sources.CHUTES }, { id: 'model_siliconflow_select', api: 'openai', type: chat_completion_sources.SILICONFLOW }, + { id: 'model_minimax_select', api: 'openai', type: chat_completion_sources.MINIMAX }, { id: 'model_electronhub_select', api: 'openai', type: chat_completion_sources.ELECTRONHUB }, { id: 'model_nanogpt_select', api: 'openai', type: chat_completion_sources.NANOGPT }, { id: 'model_deepseek_select', api: 'openai', type: chat_completion_sources.DEEPSEEK }, @@ -6628,6 +6630,32 @@ async function setApiUrlCallback({ api = null, connect = 'true', quiet = 'false' return oai_settings.siliconflow_endpoint || SILICONFLOW_ENDPOINT.GLOBAL; } + const isCurrentlyMinimax = main_api === 'openai' && oai_settings.chat_completion_source === chat_completion_sources.MINIMAX; + if (api === chat_completion_sources.MINIMAX || (!api && isCurrentlyMinimax)) { + if (!url) { + return oai_settings.minimax_endpoint || MINIMAX_ENDPOINT.GLOBAL; + } + + const permittedValues = Object.values(MINIMAX_ENDPOINT); + if (!permittedValues.includes(url)) { + !isQuiet && toastr.warning(t`Valid options are: ${permittedValues.join(', ')}`, t`MiniMax endpoint '${url}' is not a valid option.`); + return ''; + } + + if (!isCurrentlyMinimax && autoConnect) { + toastr.warning(t`MiniMax is not the currently selected API, so we cannot do an auto-connect. Consider switching to it via /api beforehand.`); + return ''; + } + + $('#minimax_endpoint').val(url).trigger('input'); + + if (autoConnect) { + $('#api_button_openai').trigger('click'); + } + + return oai_settings.minimax_endpoint || MINIMAX_ENDPOINT.GLOBAL; + } + const isCurrentlyVertexAI = main_api === 'openai' && oai_settings.chat_completion_source === chat_completion_sources.VERTEXAI; if (api === chat_completion_sources.VERTEXAI || (!api && isCurrentlyVertexAI)) { const defaultRegion = 'us-central1'; diff --git a/public/scripts/tokenizers.js b/public/scripts/tokenizers.js index 7f574e0eb..ce722b77f 100644 --- a/public/scripts/tokenizers.js +++ b/public/scripts/tokenizers.js @@ -694,6 +694,11 @@ export function getTokenizerModel() { } } + if (oai_settings.chat_completion_source == chat_completion_sources.MINIMAX) { + // MiniMax uses a proprietary tokenizer; fall back to a coarse OpenAI estimation. + return 'gpt-3.5-turbo'; + } + if (oai_settings.chat_completion_source == chat_completion_sources.WORKERS_AI && oai_settings.workers_ai_model) { const model = oai_settings.workers_ai_model.toLowerCase(); diff --git a/public/scripts/tool-calling.js b/public/scripts/tool-calling.js index 1cdbb4d2f..858908b6d 100644 --- a/public/scripts/tool-calling.js +++ b/public/scripts/tool-calling.js @@ -667,6 +667,7 @@ export class ToolManager { chat_completion_sources.SILICONFLOW, chat_completion_sources.NANOGPT, chat_completion_sources.WORKERS_AI, + chat_completion_sources.MINIMAX, ]; return supportedSources.includes(settings.chat_completion_source); } diff --git a/src/constants.js b/src/constants.js index 281228a2f..6cd43785d 100644 --- a/src/constants.js +++ b/src/constants.js @@ -209,6 +209,7 @@ export const CHAT_COMPLETION_SOURCES = { AZURE_OPENAI: 'azure_openai', ZAI: 'zai', SILICONFLOW: 'siliconflow', + MINIMAX: 'minimax', WORKERS_AI: 'workers_ai', }; @@ -557,3 +558,8 @@ export const SILICONFLOW_ENDPOINT = { GLOBAL: 'global', CN: 'cn', }; + +export const MINIMAX_ENDPOINT = { + GLOBAL: 'global', + CN: 'cn', +}; diff --git a/src/endpoints/backends/chat-completions.js b/src/endpoints/backends/chat-completions.js index dd99bfb43..df911e65f 100644 --- a/src/endpoints/backends/chat-completions.js +++ b/src/endpoints/backends/chat-completions.js @@ -18,6 +18,7 @@ import { OPENROUTER_HEADERS, VERTEX_SAFETY, SILICONFLOW_ENDPOINT, + MINIMAX_ENDPOINT, ZAI_ENDPOINT, } from '../../constants.js'; import { @@ -89,6 +90,8 @@ const API_ZAI_COMMON = 'https://api.z.ai/api/paas/v4'; const API_ZAI_CODING = 'https://api.z.ai/api/coding/paas/v4'; const API_SILICONFLOW = 'https://api.siliconflow.com/v1'; const API_SILICONFLOW_CN = 'https://api.siliconflow.cn/v1'; +const API_MINIMAX = 'https://api.minimax.io/v1'; +const API_MINIMAX_CN = 'https://api.minimaxi.com/v1'; const API_OPENROUTER = 'https://openrouter.ai/api/v1'; const API_WORKERS_AI = 'https://api.cloudflare.com/client/v4/accounts'; @@ -1552,7 +1555,87 @@ async function sendChutesRequest(request, response) { } /** - * Sends a chat completion request to Azure OpenAI. + * Sends a request to MiniMax. + * @param {express.Request} request Express request + * @param {express.Response} response Express response + */ +async function sendMinimaxRequest(request, response) { + const apiUrl = request.body.minimax_endpoint === MINIMAX_ENDPOINT.CN + ? API_MINIMAX_CN : API_MINIMAX; + const apiKey = readSecret(request.user.directories, SECRET_KEYS.MINIMAX, request.body.secret_id); + + if (!apiKey) { + console.warn('MiniMax key is missing.'); + return response.status(400).send({ error: true }); + } + + const controller = new AbortController(); + request.socket.removeAllListeners('close'); + request.socket.on('close', function () { + controller.abort(); + }); + + try { + // MiniMax does not allow consecutive messages with the same role. + // Merge them into a single message to avoid "invalid chat setting (2013)". + const messages = postProcessPrompt(request.body.messages, PROMPT_PROCESSING_TYPE.MERGE_TOOLS, getPromptNames(request)); + + let bodyParams = {}; + + if (Array.isArray(request.body.tools) && request.body.tools.length > 0) { + bodyParams['tools'] = request.body.tools; + bodyParams['tool_choice'] = request.body.tool_choice; + } + + const requestBody = { + 'messages': messages, + 'model': request.body.model, + 'temperature': request.body.temperature, + 'max_tokens': request.body.model === 'M2-her' ? Math.min(request.body.max_tokens, 2048) : request.body.max_tokens, + 'stream': request.body.stream, + 'top_p': request.body.top_p, + 'stop': request.body.stop, + ...bodyParams, + }; + + const config = { + method: 'POST', + headers: { + 'Content-Type': 'application/json', + 'Authorization': 'Bearer ' + apiKey, + }, + body: JSON.stringify(requestBody), + signal: controller.signal, + }; + + console.debug('MiniMax request:', requestBody); + + const generateResponse = await fetch(apiUrl + '/chat/completions', config); + + if (request.body.stream) { + await forwardFetchResponse(generateResponse, response); + } else { + if (!generateResponse.ok) { + const errorText = await generateResponse.text(); + console.warn('MiniMax returned error: ', errorText); + const errorJson = tryParse(errorText) ?? { error: true }; + return response.status(500).send(errorJson); + } + const generateResponseJson = await generateResponse.json(); + console.debug('MiniMax response:', generateResponseJson); + return response.send(generateResponseJson); + } + } catch (error) { + console.error('Error communicating with MiniMax: ', error); + if (!response.headersSent) { + response.send({ error: true }); + } else { + response.end(); + } + } +} + +/** * @param {express.Request} request Express request object (contains request.body with all generate_data) * @param {express.Response} response Express response object */ @@ -2096,6 +2179,7 @@ router.post('/generate', async function (request, response) { case CHAT_COMPLETION_SOURCES.AIMLAPI: return await sendAimlapiRequest(request, response); case CHAT_COMPLETION_SOURCES.XAI: return await sendXaiRequest(request, response); case CHAT_COMPLETION_SOURCES.CHUTES: return await sendChutesRequest(request, response); + case CHAT_COMPLETION_SOURCES.MINIMAX: return await sendMinimaxRequest(request, response); case CHAT_COMPLETION_SOURCES.ELECTRONHUB: return await sendElectronHubRequest(request, response); case CHAT_COMPLETION_SOURCES.AZURE_OPENAI: return await sendAzureOpenAIRequest(request, response); }