@ank1015/agents-provider-openai-azure-spec

Static Azure OpenAI provider metadata for @ank1015/agents.

This package contains the provider id, config schema, provider option/native message types, model catalog, and TypeScript module augmentation. It does not create an Azure OpenAI client or make network calls.

Install

pnpm add @ank1015/agents-provider-openai-azure-spec

Exports

• @ank1015/agents-provider-openai-azure-spec
• @ank1015/agents-provider-openai-azure-spec/config
• @ank1015/agents-provider-openai-azure-spec/models

Config

import {
  OPENAI_AZURE_PROVIDER,
  openAIAzureProviderConfigSchema,
} from '@ank1015/agents-provider-openai-azure-spec';

const config = openAIAzureProviderConfigSchema.parse({
  provider: OPENAI_AZURE_PROVIDER,
  apiKey: { type: 'env', name: 'AZURE_OPENAI_API_KEY' },
  endpoint: 'https://example.openai.azure.com',
});

Callers that already have a key from another secret source can pass it directly:

const config = openAIAzureProviderConfigSchema.parse({
  provider: OPENAI_AZURE_PROVIDER,
  apiKey: { type: 'value', value: azureOpenAIKey },
  endpoint: 'https://example.openai.azure.com',
});

Use either endpoint or baseUrl, never both.

• Without apiVersion, the runtime provider uses Azure OpenAI v1/OpenAI-compatible URLs.
• With apiVersion, the runtime provider uses the legacy AzureOpenAI SDK client.

Models

import {
  OPENAI_AZURE_MODELS,
  getOpenAIAzureModel,
} from '@ank1015/agents-provider-openai-azure-spec';

const model = getOpenAIAzureModel('gpt-5.6-terra');

The static catalog currently includes:

• gpt-5.6-sol
• gpt-5.6-terra
• gpt-5.6-luna
• gpt-5.5
• gpt-5.4-mini
• gpt-5.4-nano

GPT-5.6 Sol, Terra, Luna, and GPT-5.5 include long-context tiers: prompts over 272K input tokens use 2x input/cache rates and 1.5x output rates for the full request. GPT-5.4 mini and nano remain flat-priced. Catalog prices represent Global token rates; Azure deployment type and data zone can change the billed rate.

Azure deployment names are expected to match these static model ids.