AI Pricing, News and Router Hubs
Global App Store IAP comparison · AI router benchmarks · AI community news
Live price gap
Latest news
From X, Hacker News, Reddit and Product Hunt
- MiniMax M3 Open-Weight Model: 428B Params, 23B Activated, 1M Context & Native MultimodalJun 12
MiniMax released the M3 open-weight model with only 428B total params and 23B activated, yet it handles frontier coding, long-horizon agents, and native multimodal (text/image/video) with a 1M-token context.
Source: xAuthor: MiniMax_AIHotness: 8 - Ask HN: Show your AI coded games [June 2026]Jun 12
A discussion on Hacker News inviting developers to showcase games they coded with AI, sharing creativity and practical experiences.
Source: hackernewsAuthor: franzeHotness: 1 - Kimi 2.7 Code now available: supports Go, text, image, optimized for coding, same pricing as 2.6Jun 12
Moonshot AI launches Kimi 2.7 Code with Go support, text and image capabilities, coding optimization, at same price as 2.6.
Source: xAuthor: opencodeHotness: 196 - World of ClaudeCraft: An MMO RPG vibe coded with Fable 5Jun 12
The author created an MMO RPG called World of ClaudeCraft using vibe coding with Claude's Fable 5, showcasing AI-assisted game development.
Source: hackernewsAuthor: beatthatflightHotness: 3 - Show HN: Rubric – test what your LLM agent did, not just what it saidJun 12
Rubric is a new tool that tests LLM agents on their actual actions (tool calls, outputs) rather than just dialogue, helping developers evaluate agent performance more accurately.
Source: hackernewsAuthor: kareemrashedHotness: 1