The Azure OpenAI provider connects to Azure-hosted OpenAI deployments. It uses the same request/response format as OpenAI but with a different URL structure (/openai/deployments/{id}/...) and api-key header authentication.
Installation01
import "github.com/xraph/nexus/providers/azureopenai"Quick Start02
provider := azureopenai.New(
os.Getenv("AZURE_OPENAI_API_KEY"),
azureopenai.WithResourceName("my-resource"),
azureopenai.WithDeploymentID("gpt-4o"),
)
gw := nexus.New(
nexus.WithProvider(provider),
)Options03
| Option | Description |
azureopenai.WithResourceName(name) | Azure resource name (used to construct the base URL) |
azureopenai.WithDeploymentID(id) | Deployment ID for the model |
azureopenai.WithAPIVersion(version) | API version (default: 2024-08-01-preview) |
azureopenai.WithBaseURL(url) | Override the full base URL directly |
Capabilities04
| Capability | Supported |
| Chat | Yes |
| Streaming | Yes |
| Embeddings | Yes |
| Vision | Yes |
| Tools | Yes |
| Thinking | No |
Models05
| Model | Context | Max Output | Input Price | Output Price |
gpt-4o | 128K | 16,384 | $2.50/M | $10.00/M |
gpt-4o-mini | 128K | 16,384 | $0.15/M | $0.60/M |
text-embedding-3-small | 8,191 | — | $0.02/M | — |