OSS-first docs
These docs teach the open system first: contracts, generated surfaces, runtimes, governance, and incremental adoption. Studio shows up as the operating layer on top, not as the source of truth.
Mistral integration providing chat completions and embedding generation.
Source Definition
export const mistralIntegrationSpec = defineIntegration({
meta: {
key: 'ai-llm.mistral',
version: '1.0.0',
category: 'ai-llm',
title: 'Mistral Large Language Model',
description:
'Mistral integration providing chat completions and embedding generation.',
domain: 'ai',
owners: ['platform.ai'],
tags: ['ai', 'llm', 'embeddings'],
stability: StabilityEnum.Experimental,
},
supportedModes: ['managed', 'byok'],
capabilities: {
provides: [
{ key: 'ai.chat', version: '1.0.0' },
{ key: 'ai.embeddings', version: '1.0.0' },
],
},
configSchema: {
schema: {
type: 'object',
properties: {
model: {
type: 'string',
description:
'Default chat completion model (e.g., mistral-large-latest).',
},
embeddingModel: {
type: 'string',
description: 'Embedding model identifier.',
},
},
},
example: {
model: 'mistral-large-latest',
embeddingModel: 'mistral-embed',
},
},
secretSchema: {
schema: {
type: 'object',
required: ['apiKey'],
properties: {
apiKey: {
type: 'string',
description:
'Mistral API key with access to chat and embeddings endpoints.',
},
},
},
example: {
apiKey: 'mistral-***',
},
},
healthCheck: {
method: 'custom',
timeoutMs: 5000,
},
docsUrl: 'https://docs.mistral.ai/platform/endpoints',
constraints: {
rateLimit: {
rpm: 600,
},
},
byokSetup: {
setupInstructions:
'Generate an API key within the Mistral console and ensure the selected models are enabled for the account.',
},
});