v0.13.7
SaschaHeyer/gen-ai-livestreamv0.13.7Oct 11, 2026by tashfeenahmed
Upgrading?
AI Summary
This release expands the API with reranking and moderation endpoints, introduces a reliability floor for routing, and adds support for custom endpoint groups and new providers.
Key Highlights
- New `/v1/rerank` and `/v1/moderations` endpoints
- Minimum-reliability floor for routing to skip low-performing models/keys
- Custom endpoint groups for better organization
- Compact panel support for Windows desktop
New Features
- Rerank endpoint
- Moderation endpoint
- Reliability floor for routing
- Custom endpoint groups
- Windows compact panel
- New provider support (Inferbase, SimpleLLM, ElectronHub, Plugsky)
Full Release Notes
FreeLLMAPI v0.13.7 adds rerank and moderation endpoints, custom endpoint groups, a reliability floor for routing, and the compact panel on Windows. ## API - **`POST /v1/rerank`**: Cohere/Jina-style re-ranking over your Cohere keys or any custom endpoint that exposes `/rerank`. (#1029, #1372) - **`POST /v1/moderations`**: OpenAI-compatible moderation, routed to OpenAI, OpenRouter or NVIDIA keys with failover. (#1155) ## Routing - **Minimum-reliability floor.** Optionally skip any model or key whose observed success rate is below a set threshold. Off by default. New models are still tried, and if the floor would rule out every model it is ignored. (#1309) - **Faster routing.** The active chain, key counts and key decryption are now cached instead of being recomputed on every request. (#1318) ## Keys - **Custom endpoint groups.** Give custom endpoints a group label so they no longer all sit under one "Custom" heading, then route to a group with `custom:<model>#<group>`. (#1176, #1371) - **Health checks back off after a 429.** A key that gets rate-limited during validation is skipped by scheduled checks for an hour. (#1369, #1377) ## Desktop - **Show Compact Panel on Windows.** On Windows, clicking the tray icon opens the dashboard, so the compact panel had no way in. It now has its own item in the tray menu. Thanks @suantea for the fix, and @klaus-2 for testing it on Windows 11. (#1412, #1419) ## Providers - **Inferbase** and **SimpleLLM** chat, and **ElectronHub** embeddings. (#1455) - **Plugsky** chat and **ElectronHub** speech. (#1441) ## Updating Desktop users on v0.13.5 or later can update from inside the app: ⋯ → **Check for updates**. Docker: `ghcr.io/tashfeenahmed/freellmapi:v0.13.7`. The `latest` tag also points to this release. Existing keys and profiles are retained. This release has one database migration, which adds an optional group label to keys and runs automatically on startup. [Full changelog](https://github.com/tashfeenahmed/freellmapi/compare/v0.13.6...v0.13.7) --- ⭐ Like the free router? [Go Premium](https://freellmapi.co/?utm_source=github&utm_medium=release&utm_campaign=premium&utm_content=v0.13.7#pricing), the live signed catalog, $19/yr, cancel anytime.
Related Repositories
rasbt/LLMs-from-scratch★ 107,095
Implement a ChatGPT-like LLM in PyTorch from scratch, step by step
lyogavin/airllm★ 40,274
AirLLM 2.8T LLM inference with single 4GB GPU
NirDiamant/RAG_Techniques★ 29,967
This repository showcases various advanced techniques for Retrieval-Augmented Generation (RAG) systems. Each technique has a detailed notebook tutorial.
microsoft/OmniParser★ 25,540
A simple screen parsing tool towards pure vision based GUI agent
karpathy/nn-zero-to-hero★ 24,911
Neural Networks: Zero to Hero