Ner_Evane

🇺🇸

[lino @prettyboicashin]

la musique mdrrr ???? mais il se passe quoi là

[Nous Research @NousResearch]

The next evolution of Hermes Agent is here!

Introducing Hermes Desktop: everything you love about Hermes, now native on your machine.

First demoed in Jensen's GTC keynote, it's now in public preview.

[Shubham Sharma @HappyyPablo]

open sourcing Marlin-2B 🐟
a tiny VLM to extract structured information from videos

Marlin is finetuned for two questions devs want to ask in their videos: what is happening, and when?

Best open model in its weight class, competitive with Gemini-2.5-flash at only 2B params 🧵

[SilenceÇaPrompt @SilenceCaPrompt]

Ce chinois vient d'open-sourcer un outil qui transforme n'importe quelle donnée en HTML de niveau professionnel.

> 15 000 lignes de code en 3 jours.
> 75 skills.
> 9 formats d'export.
> Compatible Claude Code, Codex, OpenClaw et Hermes.

[Erick @ErickSky]

¿Estás cansado de pagar por APIs de WhatsApp que te limitan, te cobran por mensaje o te atan a un proveedor externo que controla todo?

¿Buscas una solución profesional, moderna y que puedas tener 100% bajo tu control sin vendor lock-in?

[OpenWA]

es un gateway de API para WhatsApp gratuito, open source y self-hosted. Te permite enviar y recibir mensajes, gestionar sesiones, configurar webhooks y automatizar flujos completos a través de una REST API limpia y bien documentada, sin pagar licencias ni depender de nadie.

No es solo "otro wrapper".

Está construido con un stack serio: NestJS 11 + TypeScript en el backend y un dashboard web en React. Todo pensado para que sea fácil de usar, mantener y escalar.

Sus características más potentes incluyen:

- Soporte real para múltiples sesiones. Puedes tener varias cuentas de WhatsApp funcionando al mismo tiempo en una misma instancia sin complicaciones.

- Webhooks con firma HMAC para recibir eventos en tiempo real de forma segura (mensajes recibidos, entregados, leídos, etc.).

- Una API completa que cubre texto, multimedia (imágenes, videos, documentos, audio), reacciones, grupos, canales de WhatsApp y gestión de etiquetas.

- Arquitectura pluggable de verdad: eliges SQLite o PostgreSQL, almacenamiento local o S3/MinIO, y caché en memoria o Redis. Puedes empezar simple y escalar cuando lo necesites sin reescribir nada.

- Funciones pensadas para uso serio: rate limiting, whitelisting por CIDR, logging de auditoría completo y soporte de proxies por sesión.

El despliegue es de los más simples que he visto. Con Docker Compose levantas todo en minutos: API + dashboard + documentación Swagger lista para usar.

Es ideal para developers que quieren construir bots, sistemas de soporte automatizado, integraciones con CRMs, herramientas internas o cualquier flujo conversacional que necesite WhatsApp de forma programática y confiable.

Y lo mejor: es open source bajo licencia MIT. Puedes revisar el código completo, adaptarlo, auditarlo o contribuir. No hay cajas negras ni dependencias ocultas.

Si valoras el control sobre tus datos y tus comunicaciones, y quieres una alternativa real y profesional a las soluciones cerradas o de pago, OpenWA es de esas herramientas que vale la pena tener en el radar.

Con casi 900 estrellas ya está demostrando que hay mucha gente buscando exactamente esto.

REPOOO👇

[How To Prompt @HowToPrompt__]

ElevenLabs just lost its moat 🤯

Someone has open-sourced a single app that replaces ElevenLabs AND WisprFlow and runs 100% locally.

→ Clone any voice from a 3 seconds of audio
→ 7 TTS engines under one roof
→ 23 languages: Arabic, Hindi, Japanese, you name it
→ Built-in MCP server so Claude Code, Cursor, and Cline can speak back to you in a voice you cloned
→ Local LLM rewrites your voice in-character before TTS
→ Pedalboard effects (reverb, pitch shift, chorus) baked in

It's built on Tauri (Rust), not Electron. Runs on MLX for Apple Silicon, CUDA, ROCm, Intel Arc, DirectML, and CPU.

ElevenLabs Creator is $99/month. WisprFlow Pro is $15/month.

Voicebox is $0. 23.4K stars on GitHub. MIT license.

[Latte @0xbisc]

Wing Chun Flow

made with GPT Image 2 + Seedance 2
#DreaminaCPP @dreamina_ai

Prompt below ↓

[Kōda @aimikoda]

Free character for anyone to use.

MEI LIN

Use it for animations, games, videos, experiments or whatever you want.
Personal or commercial use is totally fine.

[Hasan @Ubermenscchh]

Someone built a free PDF editor that does everything Adobe does, in 20MB.

RevPDF runs entirely on your device. Edit text and images, sign, redact, compress, convert to Word.. no internet, no account, no subscription.

100% Free. 100% Offline.

[Joruno @wsl8297]

跟大家安利一款背景移除利器：RMBG-2.0。

各种类型图片都能搞定，一键抠图，速度快、边缘干净，整体效果相当稳，电商主图、广告物料这类场景尤其好用。

在线体验：https://t.co/uhqV7fOYtv

模型下载：https://t.co/uRjvqR5GdB

目前模型已开源，但不支持商用，日常个��使用完全够用。

[Higgsfield AI 🧩 @higgsfield]

Evaluate the virality here:
https://t.co/wIbskfCKNc

[DRUSKI @druski]

British Actors are taking all the Roles 🎬😂

[Berk Aladag @aladagberk]

Suno ve Udio'ya her ay 10-20 dolar haraç verme dönemi kapandı.

Birileri 4GB GPU ile evde sınırsız vokal üreten open source canavarı ACE-Step UI'ı çıkardı.

Aylık limit yok, login yok, "bu tarzı yapamam" diyen sansür mekanizması yok.

Kendi bilgisayarın, kendi kuralın.

[Alexander Whedon @alex_whedon]

Introducing SubQ - a major breakthrough in LLM intelligence.

It is the first model built on a fully sub-quadratic sparse-attention architecture (SSA),

And the first frontier model with a 12 million token context window which is:

- 52x faster than FlashAttention at 1MM tokens
- Less than 5% the cost of Opus

Transformer-based LLMs waste compute by processing every possible relationship between words (standard attention).

Only a small fraction actually matter.

@subquadratic finds and focuses only on the ones that do.

That's nearly 1,000x less compute and a new way for LLMs to scale.

[Siby_dev 🇸🇳 @sibyog13]

Quelqu’un vient de lancer un éditeur de bâtiments 3D open source qui fonctionne entièrement dans votre navigateur.

Il s’appelle Pascal Editor.

https://t.co/aRBLdMjiov

[Modengsir AI @ModengSir]

做油管视频副业搞钱的都懂😭
下载油管视频、消人声、转字幕、翻译、多语言重新配音

以前手动一套折腾半小时，还频繁翻车浪费时间

挖到开源神器 Voice-Pro 直接封神✨
扔个YouTube链接全自动一条龙搞定：
✅一键下载油管视频
✅高清智能去人声
✅精准语音转录字幕
✅100+语种一键翻译
✅原声克隆自动配音

全程本地离线运行、完全免费、不用联网，隐私超安心🔒

从前半小时的活，现在2分钟搞定！

做多语言账号、搬运二次剪辑、靠油管副业变现效率直接拉满，省下超多时间专心搞收益☕

GitHub直达👉 https://t.co/LqdC4jzVPU

靠油管做副业、做自媒体变现的一定要冲！
有没有被剪辑流程搞崩溃的？还有啥好工具大家可以评论区推荐一下

[ÀBDŪLLÂH @itxabdullaa]

GPT Image 2

{
"prompt": {
"title": "Agency-Grade Brand Identity System Poster",
"trigger": "Upload a logo. From it, construct a complete, investment-worthy brand identity system poster — the kind that closes client deals and dominates Behance front pages.",

"prime_directive": {
"rule": "Every element — color, tone, shape, texture, personality — must be extracted directly from the uploaded logo.",
"enforcement": "Nothing generic. Nothing templated. Nothing borrowed. Strip the logo. Decode it. Build an entire visual universe from its DNA."
},

"format": {
"orientation": "Vertical",
"aspect_ratio": "4:5",
"layout": "Multi-column grid",
"composition": "Layered, dense, intentional — zero wasted space"
},

"sections": {

"01_brand_header": {
"label": "Open With Authority",
"elements": [
"Brand name in commanding, high-hierarchy typography",
"Brand statement — 6 words maximum, razor-sharp",
"Three soul descriptors (e.g. Raw / Futuristic / Grounded)"
]
},

"02_color_system": {
"label": "Build The Color World",
"palettes": {
"primary": "3–5 colors extracted from logo",
"secondary": "3–5 supporting colors",
"accent": "High-impact hit colors"
},
"per_color_display": [
"Wide swatch block",
"HEX code",
"Role label: foundation / emphasis / atmosphere"
],
"extras": [
"Gradient blends",
"Color-on-color pairings",
"Light mode vs dark mode behavior"
]
},

"03_typography_system": {
"label": "Establish The Type Voice",
"tiers": {
"headline": "Commanding, bold — show a punchy title example",
"subheadline": "Structured, clear — show a descriptive line example",
"body": "Readable, intentional — show a paragraph fragment example"
},
"requirement": "Hierarchy must be undeniable at a glance"
},

"04_visual_language": {
"label": "Define The Visual World",
"define": [
"Image style (editorial / industrial / cinematic / organic / etc.)",
"Lighting quality and direction",
"Texture references and material moods"
],
"visual_tiles": {
"count": "3–5 tiles",
"style": "Art-directed style previews — mood board squares from a real shoot brief"
}
},

"05_brand_applications": {
"label": "Bring The Brand To Life",
"rule": "Every mockup must feel like the same brand. Same DNA. Zero inconsistency.",
"mockups": [
{
"type": "Product Packaging",
"detail": "Dimensional, realistic render"
},
{
"type": "Website Hero",
"detail": "Full desktop viewport"
},
{
"type": "Mobile App Screen",
"detail": "One key UI moment"
},
{
"type": "Social Media Posts",
"detail": "3 formats — square, story, banner"
},
{
"type": "Business Card",
"detail": "Front and back"
},
{
"type": "Out-of-Home Ad",
"detail": "Billboard or transit panel"
}
]
},

"06_design_system": {
"label": "Show The System Working",
"components": [
"Buttons — default, hover, disabled states",
"Cards",
"Input fields",
"Navigation bar",
"Spacing scale"
],
"requirement": "Must resemble a real design system handoff document"
},

"07_iconography": {
"label": "Iconography",
"count": "6–10 icons",
"style_rule": "Same visual grammar as the logo — geometric, organic, sharp, or soft",
"consistency": "Uniform stroke weight or fill logic throughou