Awesome Opensource Email Resources
🛡️ Автоматический сборщик MTProto‑прокси с уникальной фильтрацией «Умный комбайн» для автоматического сбора, анализа и фильтрации MTProto‑прокси. Отделяет **"золотые" прокси**, маскирующиеся под Yandex, VK, Gosuslugi и другие популярные RU‑сервисы (Fake‑TLS), от обычных.
Auto-updated SOCKS5 proxy list + proxies for Telegram (прокси для телеграмм/телеграм/тг) https://mtpro.xyz/mtproto https://mtpro.xyz/webproxy
Pode is a Cross-Platform PowerShell web framework for creating REST APIs, Web Sites, and TCP/SMTP servers
Kurrier is a modern, self-hosted workspace for email, calendar, contacts and storage.
This repository holds the source code of Microsoft.Testing.Platform (MTP), a lightweight alternative to VSTest, as well as MSTest adapter and framework.
Emails at the heart of your business logic!
💌 Mail app for Nextcloud
Feature-rich single-binary file server for red teamers and developers. HTTP/S · WebDAV · FTP/SFTP · SMB · LDAP/S · NTLM hash capture · DNS/SMTP callbacks · TLS · Auth · Share links. A powerful python3 -m http.server replacement.
An Open Source Lightweight E-Mail Client
SMTP (email) server with a focus on simplicity, security, and ease of operation [mirror]
Native Telegram client for Android
基于 Cloudflare Workers 的 Telegram Web/MTProto WebSocket 代理中继,支持 bridge 鉴权、bootstrap/session 令牌、多 lane WS 复用、AES-CTR 加密、自定义帧、流控与严格 CSP。
Chatmail Rust Core library, used by Android/iOS/desktop chatmail apps, bindings and bots 📧
📨 Open-Source Email API for Developers. Cloud & Self-hosted Alternative to SES, Mailgun, SendGrid.
Create custom SMTP servers on the fly
Kurigram — actively maintained Pyrogram fork with support for Gifts, Stories, Topics, business features and the latest Telegram updates.
The Ultimate Telegram Proxy Manager Built on Rust
Multi architecture simple SMTP server (postfix null relay) host for your Docker and Kubernetes containers. Based on Debian/Ubuntu/Alpine.
DavMail POP/IMAP/SMTP/CalDAV/CardDAV/LDAP Exchange and Office 365 Gateway - Synced with main subversion repository at
Official master for the Greenmail project
The extensible Erlang SMTP client and server library.
ProtonMail IMAP/SMTP Bridge Docker container
Golang package for send email. Support keep alive connection, TLS and SSL. Easy for bulk SMTP.
Fast, simple, async php telegram api server: MadelineProto + Amp HTTP Server
Simple Golang SMTP relay/proxy server
Self-hosted email delivery platform that enables applications to send emails via HTTP APIs while handling SMTP delivery, templates, storage, security, and analytics. Fully self-hostable alternative to services like SendGrid or Mailgun.
Upyo is a simple and cross-runtime library for sending email messages using SMTP and various email providers. It works on Node.js, Deno, Bun, and edge functions.
modern typescript library for mtproto. supports browsers, node/bun, deno and more!
The first Open-Source high-performance MTA developed from the ground-up for high-volume email sending environments.
Analyze HTTP, DNS and SMTP requests and create custom responses and DNS records for your subdomain
A self-hosted Email Middleware for IMAP, SMTP, Gmail API, Graph API — built for developers
Free, self-hostable scheduling app with booking links, Google Calendar sync, SMTP notifications, and Stripe test payments.
One llama.cpp flag unlocks +33-39% decode speed for Qwen3.8-27B on consumer GPUs. The MTP head already ships inside your GGUF. Recipe, paired benchmarks, probe tool.
Нативный высокопроизводительный MTProto & Cloudflare WebSocket прокси-сервер для локального проксирования трафика Telegram через CloudFlare
Self-hosted mail server for your own domain: SMTP/IMAP/POP3, CardDAV/CalDAV, React webmail and admin console, one-command Docker Compose.
A unified, torch-native training framework for MTP and block-parallel speculative decoding.
Qwen 3.8 27B ROCmFP4 on AMD Strix Halo (Ryzen AI Max+ 395). Up to 36 tok/s via MTP Speculation, TurboQuant & Mesa RADV Wave64.
Email inbox API for AI agents. Give every agent its own real email address, receive mail as structured webhooks, and reply in-thread over REST. A complete multi-tenant mail server in Rust: inbound and outbound, DKIM/SPF/DMARC/ARC, MTA-STS, DANE, three-tier anti-spam.
GLM-5.3-Flash (NVFP4) on 2x NVIDIA DGX Spark - vLLM TP2, 262K context, MTP. World-first deploy recipe: 7 day-0 bugs found and fixed, patched sm121 image, probes and full report.
The open-source email platform — self-host on your own AWS SES, or use the cloud. Resend-compatible API.
MTProto прокси для Telegram: рабочий список мтпрокси для тг, обновляется каждый час. Free Telegram MTProxy list, verified hourly.
llama.cpp fork for significantly improved performance on Ampere (especially RTX 3090 / 3090 Ti): example: 95+ tok/s over a 100K-token generation at temperature 1 for qwen3.8, supporting full 262K ctx. Supports low memory (12gb) configs for 200k+ ctx
Open recipes, engine patches, and benchmark harnesses for LLM inference on Intel Arc Pro B60/B70 (Battlemage, Xe2). MoE 35B at 160 t/s decode / 7.5K t/s prefill single-stream, 27B at 50~ t/s decode / 1.7K t/s prefill single stream. vLLM XPU MTP unlocked. Muse Glimmer recipe added!!
Live truecolor terminal dashboard for a locally running LLM: tok/s, TTFT, GPUs, context, MTP acceptance, layer and real expert routing (llama.cpp)
Qwen3.8-Flash-Next (NVFP4) on DGX Spark in vLLM: one Spark 43.9 tok/s with our disk-backed n-gram table patch, staged gather and reduced-vocab MTP draft; TP2 SPEED 53.7; TP4 CONTEXT 9M KV pool. Launchers, patch, 40-prompt harness, KV ledger, credits. SGLang TP2 day-0 lane kept.
Qwen3.8-Flash-Next on 2× RTX 3090: with 128 GB RAM up to 4,191 tok/s prefill · 111.5 tok/s decode (131K prompt), full 256K window at 2,865 tok/s; with 64 GB RAM 3,410 tok/s prefill · 84 tok/s decode. New: opt-in uncensored mode (runtime abliteration, no new weights).
A self-hosted, open-source mail server suite in a single Node.js process — full-stack email with SMTP / IMAP / POP3, a modern three-pane webmail, an admin dashboard, multi-tenancy, DKIM, anti-spam and 2FA. Zero native dependencies, one SQLite file, up and running in minutes.
Mirai S (Qwen3.8-27B-S trellis GGUF) on a 12 GB RTX 4070: 262k q8_0 tiered cache, MTP drafting at every depth, harness-proofing, the layer; receipts.
DeepSeek-V4.1-Flash on two DGX Sparks: EXL3 routed experts, 262k context, 2M-token KV, vision, tools, MTP, bundled with Pi
Панель управления парком Telegram-прокси (telemt, MTProto): ключи, сайты-прикрытия, мониторинг · Self-hosted control panel for Telegram proxy nodes
GLM-5.2 NVFP4+AQLM on 3× DGX Spark — 380k context MTP serve stack
Native Windows port of NInfer engine for RTX 5090. Features Qwen3.8-27B with QUASAR and NInfer models, MTP/DFlash2 with vision and 262,144 context
Go port of Python's parse library: extract named, typed fields from a string using a format() template.
Qwen 3.8 27B at 48-65 tok/s on a Mac Studio M4 Max — measured oMLX native-MTP recipe, same-day A/B vs mlx-vlm+MTP sidecar, raw benchmark JSON included
Validated GLM-5.3 Flash recipe for 2x NVIDIA RTX PRO 6000 Blackwell 96GB: 262K context, EXL3/TR3, adaptive MTP, tools, and vision.
MailCraft: an open-source, clean-room mail, calendar and contacts program (Outlook-style) in pure Rust, for Microsoft 365, Outlook.com and IMAP. Early alpha.
GLM-5.3-Flash (320B MoE) at TP4 across four DGX Sparks, same day as the model drop: 36 tok/s, 1.26M-token FP8 KV pool, 262K context, MTP spec decode. First TP4 glm5_next outside B200 hardware. Full patched-image recipe + the GB10 memory study.
Telegram WEB Cloudflare Workers proxy (MTProxy relay) - 仅供学习
Qwen 3.8 27B MLX challenge — placeholder, not yet open.