Feature/add hpc ai provider

.github/README.jp.md

@@ -148,6 +148,7 @@ AIゲートウェイは、すべての他のプロバイダーとモデルに対
 | <img src="https://files.readme.io/d38a23e-small-studio-favicon.png" width=35>                                              | [AI21](https://portkey.ai/docs/welcome/integration-guides)                                    | ✅       | ✅      |
 | <img src="https://platform.stability.ai/small-logo-purple.svg" width=35>                                                   | [Stability AI](https://portkey.ai/docs/welcome/integration-guides/stability-ai)               | ✅       | ✅      |
 | <img src="https://deepinfra.com/_next/static/media/logo.4a03fd3d.svg" width=35>                                            | [DeepInfra](https://portkey.ai/docs/welcome/integration-guides)                               | ✅       | ✅      |
+| <img src="https://company.hpc-ai.com/hubfs/logos/colossal-ai_logo_vertical.svg" width=35 height=35 style="object-fit:contain" alt="HPC-AI"> | [HPC-AI](https://www.hpc-ai.com/) | ✅ | ✅ |
 | <img src="https://ollama.com/public/ollama.png" width=35>                                                                  | [Ollama](https://portkey.ai/docs/welcome/integration-guides/ollama)                           | ✅       | ✅      |
 | <img src="https://novita.ai/favicon.ico" width=35>                                                                         | Novita AI                                                                                     | ✅       | ✅      | `/chat/completions`, `/completions` |
 

README.md

@@ -260,11 +260,15 @@ Explore Gateway integrations with [45+ providers](https://portkey.wiki/gh-59) an
 | <img src="https://files.readme.io/d38a23e-small-studio-favicon.png" width=35>                                              | [AI21](https://portkey.wiki/gh-91)                                    | ✅       | ✅      |
 | <img src="https://platform.stability.ai/small-logo-purple.svg" width=35>                                                   | [Stability AI](https://portkey.wiki/gh-71)               | ✅       | ✅      |
 | <img src="https://deepinfra.com/_next/static/media/logo.4a03fd3d.svg" width=35>                                            | [DeepInfra](https://portkey.sh/gh-92)                               | ✅       | ✅      |
+| <img src="https://company.hpc-ai.com/hubfs/logos/colossal-ai_logo_vertical.svg" width=35 height=35 style="object-fit:contain" alt="HPC-AI"> | [HPC-AI](https://www.hpc-ai.com/) | ✅ | ✅ |
 | <img src="https://ollama.com/public/ollama.png" width=35>                                                                  | [Ollama](https://portkey.wiki/gh-72)                           | ✅       | ✅      |
 | <img src="https://novita.ai/favicon.ico" width=35>                                                                         | [Novita AI](https://portkey.wiki/gh-73)                              | ✅       | ✅      | `/chat/completions`, `/completions` |
 
 
 > [View the complete list of 200+ supported models here](https://portkey.wiki/gh-74)
+
+**HPC-AI:** Use `provider: "hpc-ai"` with your HPC-AI API key. The gateway targets `https://api.hpc-ai.com/inference/v1` (OpenAI-compatible chat completions). Override the base URL with `x-portkey-custom-host` or `custom_host` in config if required. Example models: `minimax/minimax-m2.5`, `moonshotai/kimi-k2.5`. For local integration tests, optional env vars: `HPC_AI_API_KEY`, `HPC_AI_BASE_URL` (set these in your environment or `.env` file as appropriate).
+
 <br>
 
 <br>

src/data/models.json

@@ -26,6 +26,22 @@
       },
       "name": "Axon Code"
     },
+    {
+      "id": "minimax/minimax-m2.5",
+      "object": "model",
+      "provider": {
+        "id": "hpc-ai"
+      },
+      "name": "MiniMax M2.5"
+    },
+    {
+      "id": "moonshotai/kimi-k2.5",
+      "object": "model",
+      "provider": {
+        "id": "hpc-ai"
+      },
+      "name": "Kimi K2.5"
+    },
     {
       "id": "meta-llama/llama-3.1-70b-instruct/fp-8",
       "object": "model",

src/globals.ts

@@ -97,6 +97,7 @@ export const LEPTON: string = 'lepton';
 export const KLUSTER_AI: string = 'kluster-ai';
 export const NSCALE: string = 'nscale';
 export const HYPERBOLIC: string = 'hyperbolic';
+export const HPC_AI: string = 'hpc-ai';
 export const BYTEZ: string = 'bytez';
 export const FEATHERLESS_AI: string = 'featherless-ai';
 export const KRUTRIM: string = 'krutrim';
@@ -173,6 +174,7 @@ export const VALID_PROVIDERS = [
   KLUSTER_AI,
   NSCALE,
   HYPERBOLIC,
+  HPC_AI,
   BYTEZ,
   FEATHERLESS_AI,
   KRUTRIM,

src/providers/hpc-ai/api.ts

@@ -0,0 +1,29 @@
+import { ProviderAPIConfig } from '../types';
+
+const DEFAULT_HPC_AI_BASE_URL = 'https://api.hpc-ai.com/inference/v1';
+
+const HpcAiApiConfig: ProviderAPIConfig = {
+  getBaseURL: ({ providerOptions }) => {
+    const fromEnv =
+      typeof process !== 'undefined' && process.env?.HPC_AI_BASE_URL
+        ? process.env.HPC_AI_BASE_URL
+        : '';
+    return providerOptions.customHost || fromEnv || DEFAULT_HPC_AI_BASE_URL;
+  },
+  headers: ({ providerOptions }) => {
+    return {
+      Authorization: `Bearer ${providerOptions.apiKey}`,
+    };
+  },
+  getEndpoint: ({ fn }) => {
+    switch (fn) {
+      case 'chatComplete':
+      case 'stream-chatComplete':
+        return '/chat/completions';
+      default:
+        return '';
+    }
+  },
+};
+
+export default HpcAiApiConfig;

src/providers/hpc-ai/chatComplete.ts

@@ -0,0 +1,230 @@
+import { HPC_AI } from '../../globals';
+import { Params } from '../../types/requestBody';
+import {
+  ChatCompletionResponse,
+  ErrorResponse,
+  ProviderConfig,
+} from '../types';
+import {
+  generateErrorResponse,
+  generateInvalidProviderResponseError,
+} from '../utils';
+
+export const HpcAiChatCompleteConfig: ProviderConfig = {
+  model: {
+    param: 'model',
+    required: true,
+    default: 'minimax/minimax-m2.5',
+  },
+  messages: {
+    param: 'messages',
+    required: true,
+    default: [],
+    transform: (params: Params) => {
+      return params.messages?.map((message) => {
+        if (message.role === 'developer') return { ...message, role: 'system' };
+        return message;
+      });
+    },
+  },
+  frequency_penalty: {
+    param: 'frequency_penalty',
+    default: 0,
+    min: -2,
+    max: 2,
+  },
+  max_tokens: {
+    param: 'max_tokens',
+    default: 100,
+    min: 1,
+  },
+  max_completion_tokens: {
+    param: 'max_tokens',
+    default: 100,
+    min: 1,
+  },
+  n: {
+    param: 'n',
+    default: 1,
+    min: 1,
+    max: 1,
+  },
+  presence_penalty: {
+    param: 'presence_penalty',
+    min: -2,
+    max: 2,
+    default: 0,
+  },
+  temperature: {
+    param: 'temperature',
+    default: 1,
+    min: 0,
+    max: 2,
+  },
+  top_p: {
+    param: 'top_p',
+    default: 1,
+    min: 0,
+    max: 1,
+  },
+  stop: {
+    param: 'stop',
+    default: null,
+  },
+  stream: {
+    param: 'stream',
+    default: false,
+  },
+};
+
+interface HpcAiChatCompleteResponse extends ChatCompletionResponse {
+  id: string;
+  object: string;
+  created: number;
+  model: string;
+  usage: {
+    prompt_tokens: number;
+    completion_tokens: number;
+    total_tokens: number;
+  };
+}
+
+export interface HpcAiDetailErrorResponse {
+  detail: {
+    loc: string[];
+    msg: string;
+    type: string;
+  }[];
+}
+
+interface HpcAiStreamChunk {
+  id: string;
+  object: string;
+  created: number;
+  model: string;
+  choices: {
+    delta: {
+      role?: string | null;
+      content?: string;
+    };
+    index: number;
+    finish_reason: string | null;
+  }[];
+  usage?: {
+    prompt_tokens: number;
+    completion_tokens: number;
+    total_tokens: number;
+  };
+}
+
+export const HpcAiChatCompleteResponseTransform: (
+  response:
+    | HpcAiChatCompleteResponse
+    | HpcAiDetailErrorResponse
+    | ErrorResponse,
+  responseStatus: number
+) => ChatCompletionResponse | ErrorResponse = (response, responseStatus) => {
+  if (responseStatus !== 200 && 'error' in response && response.error) {
+    return generateErrorResponse(
+      {
+        ...response.error,
+      },
+      HPC_AI
+    );
+  }
+
+  if (
+    'detail' in response &&
+    responseStatus !== 200 &&
+    Array.isArray(response.detail) &&
+    response.detail.length
+  ) {
+    let firstError: Record<string, any> | undefined;
+    let errorField: string | null = null;
+    let errorMessage: string | undefined;
+    let errorType: string | null = null;
+
+    [firstError] = response.detail;
+    errorField = firstError?.loc?.join('.') ?? '';
+    errorMessage = firstError.msg;
+    errorType = firstError.type;
+
+    return generateErrorResponse(
+      {
+        message: `${errorField ? `${errorField}: ` : ''}${errorMessage}`,
+        type: errorType,
+        param: null,
+        code: null,
+      },
+      HPC_AI
+    );
+  }
+
+  if ('choices' in response) {
+    return {
+      id: response.id,
+      object: response.object,
+      created: response.created,
+      model: response.model,
+      provider: HPC_AI,
+      choices: response.choices.map((c) => ({
+        index: c.index,
+        message: {
+          role: c.message.role,
+          content: c.message.content,
+        },
+        finish_reason: c.finish_reason,
+      })),
+      usage: {
+        prompt_tokens: response.usage?.prompt_tokens,
+        completion_tokens: response.usage?.completion_tokens,
+        total_tokens: response.usage?.total_tokens,
+      },
+    };
+  }
+
+  return generateInvalidProviderResponseError(response, HPC_AI);
+};
+
+export const HpcAiChatCompleteStreamChunkTransform: (
+  response: string
+) => string = (responseChunk) => {
+  if (
+    responseChunk.match(
+      /^:\s*ping\s*-\s*\d{4}-\d{2}-\d{2}\s\d{2}:\d{2}:\d{2}\.\d{6}\+\d{2}:\d{2}$/
+    )
+  ) {
+    return '';
+  }
+
+  let chunk = responseChunk.trim();
+  chunk = chunk.replace(/^data: /, '');
+  chunk = chunk.trim();
+  if (chunk === '[DONE]') {
+    return `data: ${chunk}\n\n`;
+  }
+  const parsedChunk: HpcAiStreamChunk = JSON.parse(chunk);
+  return (
+    `data: ${JSON.stringify({
+      id: parsedChunk.id,
+      object: parsedChunk.object,
+      created: parsedChunk.created,
+      model: parsedChunk.model,
+      provider: HPC_AI,
+      choices: [
+        {
+          index: parsedChunk.choices[0].index,
+          delta: parsedChunk.choices[0].delta,
+          finish_reason: parsedChunk.choices[0].finish_reason,
+        },
+      ],
+      usage: parsedChunk.usage
+        ? {
+            prompt_tokens: parsedChunk.usage.prompt_tokens,
+            completion_tokens: parsedChunk.usage.completion_tokens,
+            total_tokens: parsedChunk.usage.total_tokens,
+          }
+        : undefined,
+    })}` + '\n\n'
+  );
+};

src/providers/hpc-ai/index.ts

@@ -0,0 +1,18 @@
+import { ProviderConfigs } from '../types';
+import HpcAiApiConfig from './api';
+import {
+  HpcAiChatCompleteConfig,
+  HpcAiChatCompleteResponseTransform,
+  HpcAiChatCompleteStreamChunkTransform,
+} from './chatComplete';
+
+const HpcAiConfig: ProviderConfigs = {
+  chatComplete: HpcAiChatCompleteConfig,
+  api: HpcAiApiConfig,
+  responseTransforms: {
+    chatComplete: HpcAiChatCompleteResponseTransform,
+    'stream-chatComplete': HpcAiChatCompleteStreamChunkTransform,
+  },
+};
+
+export default HpcAiConfig;

src/providers/index.ts

@@ -60,6 +60,7 @@ import LeptonConfig from './lepton';
 import KlusterAIConfig from './kluster-ai';
 import NscaleConfig from './nscale';
 import HyperbolicConfig from './hyperbolic';
+import HpcAiConfig from './hpc-ai';
 import { FeatherlessAIConfig } from './featherless-ai';
 import KrutrimConfig from './krutrim';
 import AI302Config from './302ai';
@@ -133,6 +134,7 @@ const Providers: { [key: string]: ProviderConfigs } = {
   'kluster-ai': KlusterAIConfig,
   nscale: NscaleConfig,
   hyperbolic: HyperbolicConfig,
+  'hpc-ai': HpcAiConfig,
   bytez: BytezConfig,
   'featherless-ai': FeatherlessAIConfig,
   krutrim: KrutrimConfig,

src/public/index.html

@@ -1109,6 +1109,7 @@ <h3>Select Provider</h3>
             <option value="openai">OpenAI</option>
             <option value="anthropic">Anthropic</option>
             <option value="groq">Groq</option>
+            <option value="hpc-ai">HPC-AI</option>
             <option value="bedrock">Bedrock</option>
             <option value="azure-openai">Azure OpenAI</option>
             <option value="cohere">Cohere</option>
@@ -1460,6 +1461,7 @@ <h3>Enter API Key</h3>
         "openai": "gpt-4o-mini",
         "anthropic": "claude-3-5-sonnet-20240620",
         "groq": "llama3-70b-8192",
+        "hpc-ai": "minimax/minimax-m2.5",
         "bedrock": "anthropic.claude-3-sonnet-20240229-v1:0",
         "azure-openai": "gpt-4o-mini",
         "cohere": "command-r-plus",
@@ -1474,6 +1476,7 @@ <h3>Enter API Key</h3>
         "openai": "https://portkey.ai/docs/integrations/llms/openai",
         "anthropic": "https://portkey.ai/docs/integrations/llms/anthropic",
         "groq": "https://portkey.ai/docs/integrations/llms/groq",
+        "hpc-ai": "https://portkey.ai/docs/integrations/llms/hpc-ai",
         "bedrock": "https://portkey.ai/docs/integrations/llms/aws-bedrock",
         "azure-openai": "https://portkey.ai/docs/integrations/llms/azure-openai",
         "cohere": "https://portkey.ai/docs/integrations/llms/cohere",

src/tests/resources/testVariables.ts

@@ -62,6 +62,10 @@ const testVariables: TestVariables = {
     apiKey: process.env.DEEPINFRA_API_KEY,
     chatCompletions: { model: 'meta-llama/Meta-Llama-3-8B-Instruct' },
   },
+  'hpc-ai': {
+    apiKey: process.env.HPC_AI_API_KEY,
+    chatCompletions: { model: 'minimax/minimax-m2.5' },
+  },
   'stability-ai': {
     apiKey: process.env.STABILITY_AI_API_KEY,
     chatCompletions: { model: '' },