Endpoint reference

The canonical list of HTTP endpoints the Mira API actually serves today. Anything not on this page is not live — if you find an SDK example that calls a path not listed here, mail support@vmira.ai and we'll get the docs corrected.

Current rates

Current ratesUSD per 1M tokens
Single source of truth.Open the full pricing page →

Base URL

https://api.vmira.ai

Authentication

Authorization: Bearer sk-mira-YOUR_API_KEY

Get a key at platform.vmira.ai/api-keys — full details in Authentication.

Chat completions

The Mira conversational endpoint, drop-in compatible with the OpenAI Chat Completions API. Full reference on the chat completions page.

MethodPathPurpose
POST/v1/chat/completionsSend a conversation and receive a model-generated reply (supports streaming via SSE).
GET/v1/modelsList the models available to your key.

Mira Image

Image generation and editing. Capabilities and example code on Mira Image.

MethodPathPurpose
POST/v1/images/generateText → image.
POST/v1/images/editImage → image with a reference.

Mira Voice

Speech-to-text and text-to-speech. Capabilities and examples on Mira Voice.

MethodPathPurpose
POST/v1/audio/transcribeSpeech-to-text.
POST/v1/audio/speechText-to-speech (synthesize).
GET/v1/audio/voicesList the available voices.

Mira Realtime

Duplex voice channel over WebSocket. Protocol details on Mira Realtime.

MethodPathPurpose
WSS/v1/realtimeDuplex voice WebSocket.
GET/v1/realtime/preflightNegotiate session capabilities before connecting.

Mira Vision

Grounded object detection — given an image and a prompt, returns bounding boxes. Free-form Q&A on images goes through chat completions instead (Mira accepts image_url or base64 in messages[].content).

MethodPathPurpose
POST/v1/vision/locateGrounded object detection.

Account and management

Endpoints that read your account state. All require authentication.

MethodPathPurpose
GET/v1/billing/balanceCurrent balance and spending summary.
GET/v1/billing/transactionsRecent charges and top-ups.
GET/v1/billing/pricingThe live per-token price table for every model (powers the page you're reading).
GET/v1/usageAggregate usage stats for the account.

Next steps