Models & repos · most watched this week
Trending models
- convaiinnovations/laya
A calibrated decision model in the style of Jev. Apache 2.0 license.
Source: Hugging Face
Repos of the week
- paperclipai/paperclip
Open source app to manage agents at work.
- vectorize-io/hindsight
Hindsight: agent memory that learns.
- cloudflare/security-audit-skill
Cloudflare's skill for coding agents: multi-phase security audits with independently verified, machine-readable findings.
- Tencent/WeKnora
Tencent's open source platform that turns documents into a queryable RAG, an agent and a self-maintaining wiki.
- davila7/claude-code-templates
A command-line tool to configure and monitor Claude Code.
- stablyai/orca
An environment for running a fleet of coding agents in parallel with your own subscription. Desktop, mobile and remote.
- vercel/next.js
The React framework.
- HKUDS/CLI-Anything
CLI-Anything: making all software agent-native.
- pytorch/pytorch
Tensors and dynamic neural networks in Python with strong GPU acceleration.
- anthropics/claude-code
Anthropic's coding agent that works from the terminal.
- affaan-m/ECC
A system to tune coding agents: skills, memory and security for Claude Code, Codex, Opencode and Cursor.
Source: GitHub Trending
New: agents, skills and MCP
- jev-chat/jev-chat-jarvis
A phone copilot that reads chats in WeChat, QQ, X or Feishu and suggests replies. Sending stays with the user.
- yetone/magpie
Every agent's model in one place, from the menu bar: Codex on DeepSeek, Claude Code on Kimi.
- mikehasa/golive-skill
Takes an agent-built product live (hosting, database, domain, email and payments) on your own accounts. An open source skill and zero-dependency Node CLI, with no account or telemetry.
- nokia-applied-research/AnyJev
Turns any LLM into a Jev-style decision model with real probabilities and no retraining.
- yukitorido/short-video-generator-AI
AI video processing pipeline that generates vertical shorts using LLMs, Whisper transcription, highlight detection and automated editing.
- jev-chat/jev-chat-windows
A WeChat helper for Windows: local OCR reads the messages, Jev reads the intent and suggests three replies. Sending is always manual.
- dzhng/jevgrep
Find code by asking what it does. A CLI for coding agents that uses Jev to discover relevant files and source context.
- alexgreensh/anidoodle
Hand-drawn art and animation, written as code: illustrations, loops, stickers and scored films in nine styles, identical on every render.
- kydlikebtc/awesome-jev
1207 public resources on Jev, TypeSafe AI's System One decision model, indexed by decision pattern, with sources and dated link checks. Runtime and performance are not independently tested.
- amitshekhariitbhu/ai-system-design
A step-by-step guide to designing AI systems built on LLMs, RAG and AI agents.
- jev-chat/jev-chat-jarvis-mac
A floating WeChat window for macOS: reads the screen, a local model detects intent and risk, and suggests replies. Read-only.
- Alex314618-create/JevRev
A workflow that combines an LLM with Jev.
Source: GitHub
Hugging Face sorted by trend; GitHub Trending by stars this week; new repos are from the last 7 days with the topics ai-agents, mcp, agent-skills, claude-code or llm.
Price per million tokens, by lab
Sep 27, 17:03 UTC| Model | Input | Output |
|---|---|---|
| Claude Fable 5.1 | 10.00 | 50.00 |
| Claude Opus 5.5 | 4.00 | 20.00 |
| Claude Sonnet 5 | 2.00 | 10.00 |
| Model | Input | Output |
|---|---|---|
| GPT-6 Astra | 10.00 | 50.00 |
| GPT-6 Sol | 2.00 | 10.00 |
| GPT-6 Luna | 0.10 | 0.50 |
| Model | Input | Output |
|---|---|---|
| Gemini 3.8 Flash | 0.75 | 3.75 |
| Gemini 3.5 Flash Lite | 0.30 | 2.50 |
| Gemma 4 31B | 0.09 | 0.34 |
| Model | Input | Output |
|---|---|---|
| Grok 4.7 | 1.60 | 4.80 |
| Grok Build 0.1 | 1.00 | 2.00 |
| Model | Input | Output |
|---|---|---|
| DeepSeek V4 Pro | 0.25 | 3.50 |
| DeepSeek V4.1 Flash | 0.04 | 0.29 |
| Model | Input | Output |
|---|---|---|
| Qwen3.8 Max Prime | 4.00 | 12.00 |
| Qwen3.8 Max | 2.00 | 6.00 |
| Qwen3.8 Flash | 0.15 | 0.47 |
| Model | Input | Output |
|---|---|---|
| Kimi K3 | 3.00 | 15.00 |
| Kimi K2.7 Code | 0.66 | 3.30 |
| Model | Input | Output |
|---|---|---|
| GLM-5.3 Prime | 2.80 | 8.80 |
| GLM-5.3 | 1.40 | 4.40 |
| GLM-5.3 Flash | 0.05 | 0.14 |
USD per million tokens, input and output, from OpenRouter. Each lab's first row is its top model.