v0.9.82 v3.8.17

Name: OmniRoute
Author: cheaper-inference

high risk

cheaper-inference/OmniRouteJun 9, 2026 → Jun 9, 20260 days between releases

What's New

Adds AI providers (LMArena, ZenMux, Gemini Business) and auto-combos API

New Features

LMArena provider with streaming chat completions
ZenMux provider with OpenAI-compatible endpoint
Gemini Business provider (Google Workspace)
Auto-combos API endpoint (GET /api/combos/auto)
Connection pagination and health filtering

Improvements

Free model quota display on provider pages
MCP-plugin surface extended to match
Per-prefix API format selection (OpenAI/Anthropic/Gemini)
Debug logging via environment variable
Free-label normalizer across providers

Bug Fixes

Fixed Vertex AI tool calls
Fixed Claude Code beta flags
Fixed Codex CLI model catalog refresh
Fixed Codex CLI version header
Fixed translator bugs

Upgrade Notes

The release notes do not specify migration steps. Verify configuration compatibility as the project scope appears to have shifted from OSINT tools to AI providers.

The Two Releases

v0.9.82

ShadowBroker v0.9.82

Jun 9, 2026· 5 months ago

Introduces a Telegram OSINT map layer, Osiris-derived intel ports, and OpenClaw agent parity, alongside a GFW fishing layer and map UX fixes.

Full release notes

v3.8.17

Jun 9, 2026· 5 months ago

Adds support for three new providers (LMArena, ZenMux, Gemini Business) and the auto-combos API. The release also fixes translator bugs, catalog issues, and command-code integration.

Full release notes