Skip to main content
Glama

ACE MCP Server

mistral.tsโ€ข4.49 kB
import axios from 'axios'; import { LLMProvider, Message, ChatOptions, MistralConfig } from './provider'; import { LLMProviderError } from '../utils/errors'; import { logger } from '../utils/logger'; export class MistralProvider implements LLMProvider { public readonly name = 'mistral'; private client: any; private config: MistralConfig; constructor(config: MistralConfig) { this.config = { model: 'mistral-large-latest', // Default to Large model baseUrl: 'https://api.mistral.ai/v1', timeout: 60000, maxRetries: 3, ...config }; this.client = axios.create({ baseURL: this.config.baseUrl, timeout: this.config.timeout, headers: { 'Content-Type': 'application/json', 'Authorization': `Bearer ${this.config.apiKey}` } }); logger.info('Mistral provider initialized', { model: this.config.model, baseUrl: this.config.baseUrl }); } async chat(messages: Message[], options?: ChatOptions): Promise<string> { try { logger.debug('Mistral chat request', { messageCount: messages.length, model: options?.model || this.config.model }); const response = await this.client.post('/chat/completions', { model: options?.model || this.config.model, messages: messages.map(m => ({ role: m.role, content: m.content })), temperature: options?.temperature, max_tokens: options?.maxTokens, stream: false }); const content = response.data.choices[0]?.message?.content; if (!content) { throw new Error('Empty response from Mistral'); } logger.debug('Mistral chat completed', { model: response.data.model, usage: response.data.usage, finishReason: response.data.choices[0]?.finish_reason }); return content; } catch (error) { logger.error('Mistral chat failed', { error }); if (error && typeof error === 'object' && 'response' in error) { const axiosError = error as any; const message = axiosError.response?.data?.error?.message || axiosError.message; throw new LLMProviderError( `Mistral chat failed: ${message}`, 'mistral', axiosError ); } throw new LLMProviderError( `Mistral chat failed: ${(error as Error).message}`, 'mistral', error as Error ); } } async embed(text: string): Promise<number[]> { try { logger.debug('Mistral embedding request', { textLength: text.length }); // Mistral uses mistral-embed model for embeddings const response = await this.client.post('/embeddings', { model: 'mistral-embed', input: text }); const embedding = response.data.data[0]?.embedding; if (!embedding) { throw new Error('Empty embedding from Mistral'); } logger.debug('Mistral embedding completed', { dimensions: embedding.length }); return embedding; } catch (error) { logger.error('Mistral embedding failed', { error }); if (error && typeof error === 'object' && 'response' in error) { const axiosError = error as any; const message = axiosError.response?.data?.error?.message || axiosError.message; throw new LLMProviderError( `Mistral embedding failed: ${message}`, 'mistral', axiosError ); } throw new LLMProviderError( `Mistral embedding failed: ${(error as Error).message}`, 'mistral', error as Error ); } } async listModels(): Promise<string[]> { try { logger.debug('Mistral list models request'); const response = await this.client.get('/models'); const models = response.data.data.map((m: any) => m.id); logger.debug('Mistral list models completed', { modelCount: models.length }); return models; } catch (error) { logger.error('Mistral list models failed', { error }); // Return known models if API fails const knownModels = [ 'mistral-large-latest', 'mistral-medium-latest', 'mistral-small-latest', 'mixtral-8x7b-instruct', 'mixtral-8x22b-instruct' ]; logger.info('Returning known Mistral models', { models: knownModels }); return knownModels; } } }

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/Angry-Robot-Deals/ace-mcp'

If you have feedback or need assistance with the MCP directory API, please join our Discord server