Starstruk Logo MarkStarstruk

v0.2.0

dmtrKovalenko/odiffv0.2.0Sep 2, 2026by github-actions[bot]

AI Summary

Adds OpenAI Fast grouping strategy, model-specific reasoning effort configuration, and native Kimi Responses API forwarding support to the AI API Gateway platform.

Key Highlights

  • Support for OpenAI Fast grouping with free requests billed at Standard price
  • Configurable OpenAI reasoning effort mapping per model
  • Native Kimi OpenAI Responses API forwarding support
  • Support for Claude Fable 5.1

New Features

  • OpenAI Fast grouping strategy
  • Reasoning effort mapping
  • Kimi native Responses API forwarding
  • Claude Fable 5.1 support
  • Scheduled automation requests without call ID

Full Release Notes

> AI API Gateway Platform - 将 AI 订阅配额分发和管理

新增 OpenAI Fast 分组策略、按模型配置 reasoning effort，以及 Kimi 原生 Responses API 转发支持。

## 新增功能

- 分组支持配置 OpenAI Fast，并支持将免费 Fast 请求按 Standard 价格计费
- 支持按模型设置 OpenAI reasoning effort 映射，并可配置超限时拒绝或降级
- 支持转发 Kimi 原生 OpenAI Responses API
- 支持 Claude Fable 5.1
- 支持没有 call ID 的定时自动化启动请求

## 优化改进

- 优化分组模型定价弹窗布局
- 保留调度器快照中的 OpenAI passthrough 配置
- 改进 OpenAI API Key 对话缓存身份处理

## Bug 修复

- 修复 WebSocket 在收到 terminal event 前关闭时的处理问题
- 修复模型级冷却导致 `model_not_found` 被错误转换为 `429` 的问题
- 修复未启用 server-side-fallback beta 时仍透传 Anthropic fallback 的问题



---

## 📥 Installation

**Docker:**
```bash
# Docker Hub
docker pull weishaw/sub2api:0.2.0

# GitHub Container Registry
docker pull ghcr.io/wei-shaw/sub2api:0.2.0
```

**One-line install (Linux):**
```bash
curl -sSL https://raw.githubusercontent.com/Wei-Shaw/sub2api/main/deploy/install.sh | sudo bash
```

**Manual download:**
Download the appropriate archive for your platform from the assets below.

## 📚 Documentation

- [GitHub Repository](https://github.com/Wei-Shaw/sub2api)
- [Installation Guide](https://github.com/Wei-Shaw/sub2api/blob/main/deploy/README.md)