Orqen Docs

API Reference

API Reference

Base URL: https://api.orqen.app

Use your Orqen API key in the Authorization header for both chat completions and account APIs: Authorization: Bearer sk-orq-YOUR_KEY. The browser dashboard may use a session internally, but scripts and agents should use Orqen API keys.

Chat

GET
/v1/models

List models available from your connected providers (OpenAI-compatible shape)

POST
/v1/chat/completions

Send a chat completion request with payload optimization

POST
/v1/messages

Anthropic Messages API — use with Anthropic SDK base_url=https://api.orqen.app

Bedrock compatibility

POST
/model/{model_id}/converse

Bedrock Converse API (boto3 endpoint_url=https://api.orqen.app)

POST
/model/{model_id}/converse-stream

Bedrock Converse streaming

POST
/model/{model_id}/invoke

Bedrock InvokeModel (legacy)

Proxy endpoint families

POST
/v1/completions

Text completions via upstream proxy

POST
/v1/embeddings

Embeddings via upstream proxy

POST
/v1/responses

Responses API via upstream proxy

POST
/v1/images/generations

Image generation

POST
/v1/images/edits

Image editing

POST
/v1/audio/transcriptions

Audio transcription

POST
/v1/audio/speech

Text to speech

POST
/v1/moderations

Moderation

POST
/v1/rerank

Rerank

POST
/v1/ocr

OCR

POST
/v1/videos/generations

Video generation

Account

GET
/v1/account/me

Return the current account profile

DELETE
/v1/account/me

Permanently delete the current account and associated data

POST
/v1/account/register

Dashboard signup provisioning endpoint

API Keys

GET
/v1/account/keys

List all API keys for your account

POST
/v1/account/keys

Create a new API key

DELETE
/v1/account/keys/{key_id}

Revoke an API key

GET
/v1/keys

Short alias for listing API keys

POST
/v1/keys

Short alias for creating API keys

DELETE
/v1/keys/{key_id}

Short alias for revoking API keys

Providers

GET
/v1/account/providers

List connected provider keys

PUT
/v1/account/providers

Add or update a provider key

DELETE
/v1/account/providers/{provider}

Remove a provider key

Model routing

GET
/v1/account/routing/preferences

Get routing mode and preferences

PUT
/v1/account/routing/preferences

Update routing mode

GET
/v1/account/routing/models

List models available from connected providers

GET
/v1/account/routing/models/status

Available models with enabled state and performance data

POST
/v1/account/routing/models/toggle

Enable or disable a model for automatic routing

GET
/v1/account/routing/insights

Best recall, fastest, and most-used routing insights

GET
/v1/account/routing/calibration

Measured K policy and recall calibration buckets

GET
/v1/account/routing/traces

Recent routing decisions with intent, K reason, and top candidates

Usage, requests, and billing

GET
/v1/account/usage/summary

Current month token savings, cost, and provider savings

GET
/v1/account/usage/daily

Daily token savings for the last 30 days

GET
/v1/account/requests

Request log (paginated)

GET
/v1/account/requests/{request_id}

Detailed request log entry

GET
/v1/account/sessions

Conversation sessions (grouped by X-Orqen-Session-Id)

GET
/v1/account/sessions/{session_id}/requests

Requests within a session

GET
/v1/account/usage/analytics/daily

Daily success rate and latency percentiles

GET
/v1/account/usage/analytics/savings

Tokens saved by model and provider

GET
/v1/account/usage/analytics/latency

Overall p50/p95 latency for the period

GET
/v1/usage/analytics/daily

API-key alias for daily analytics

GET
/v1/usage/analytics/savings

API-key alias for savings breakdown

GET
/v1/usage/analytics/latency

API-key alias for latency summary

GET
/v1/account/billing

Subscription tier and projected charge

GET
/v1/usage/summary

Short alias for current month usage

GET
/v1/usage/daily

Short alias for daily usage

GET
/v1/requests

Short alias for recent requests

Optimization stats

GET
/v1/account/compression/stats

Context compression savings and technique breakdown

GET
/v1/account/strategy/stats

Optimization strategy usage (pruning, compression, recovery)

Billing (dashboard session)

POST
/v1/billing/create-checkout

Start Stripe Checkout to upgrade to Pro — requires Supabase browser session, not API key

POST
/v1/billing/customer-portal

Open Stripe customer portal — dashboard session auth

POST
/v1/billing/sync-checkout

Sync subscription after Checkout redirect — dashboard session auth

POST
/v1/billing/sync-subscription

Refresh subscription state from Stripe — dashboard session auth

Tool health

GET
/v1/account/tools/health

Description quality and routing accuracy per tool

GET
/v1/account/tools/recall-alerts

Tools that were needed but missing from forwarded sets

GET
/v1/account/tools/suggestions

Description and example suggestions from hard-negative collisions

Common response headers

x-orqen-tools-inputNumber of tools in the original request
x-orqen-tools-outputNumber of tools forwarded to the LLM
x-orqen-prune-ratiooutput/input (e.g. 8/32)
x-orqen-routingTool routing mode when set: semantic, keyword (omitted if no tools or no routing)
x-orqen-ctx-tokens-inEstimated context tokens before compression (when compression ran)
x-orqen-ctx-tokens-outEstimated context tokens after compression
x-orqen-ctx-tokens-savedContext tokens saved by compression (in − out)
X-Request-IDUnique request ID for tracing

Request logs may show additional routing values (passthrough, heuristic, intent_guided, passthrough_small_toolset, limit, and composite suffixes such as _ab_ctrl). See the payload optimization docs for pruning modes.