AI Agents3 weeks ago
SEAOTTER — Always-on Hermes Agent. No servers to run.
Create and run always-on Hermes agents without managing servers or infrastructure.
@RyanRayMartin · X
The full gallery
Tech stack
20 projects
Create and run always-on Hermes agents without managing servers or infrastructure.
@RyanRayMartin · X
Convert OpenAPI specs and databases into MCP tools without coding.
mhmt_dmr · HN
MCP Gateway – Turn existing APIs and databases into MCP servers
Publish private, public, or shareable iPhone widgets powered by your content, automations, and APIs. No app required.
@talkravarusic · X
Building Glance, it’s lets agents step out of the chat box and into your iPhone Home Screen
Access open AI models via OpenAI-compatible API using crypto payments without account creation.
@vagaziner · X
AI access still starts with a form. email card KYC an account you will forget a 5% tax sitting on top of the model bill @iamzhe built the opposite of that stack. is one OpenAI-shaped endpoint. Paste the base URL into Claude Code, Cursor, Codex, OpenCode, whatever you already use. Drop in a key. Pay with USDC, SOL, or whatever Solana token is sitting in the wallet. Fifty cents is enough to start. No signup. No passport photo. Credits do not expire. They also do not withdraw, on purpose, so the product never has to become a bank. Two hundred thirty nine models behind one key. Llama, DeepSeek, Qwen, Gemini, GLM, the weekly flavor of the month. If you do not want to pick, send minirouter/auto and let the gateway choose. If a provider dies mid-call, it fails over and tells you who actually answered and what it cost. OpenRouter takes five percent. Minirouter takes zero on the deposit and zero markup on open-weight routes. Ten dollars in is ten dollars of inference at
Infrastructure service for spinning up isolated browser sessions controlled via API.
@ignacyrusz · X
Test the Jovethra OpenAI-compatible API in a browser playground without signing up.
Belius · HN
Jovethra – prepaid OpenAI-compatible API with hard usage limits
Draft, publish, and schedule Substack posts via Python SDK, CLI, or MCP without the browser editor.
bigmuzzy · HN
The Posting Substack API
Reduce AI API costs using smart caching, model routing, and prompt compression.
@PrathamGup34482 · X
Day 1 of building in public. TokenSave: a proxy that automatically reduces AI API costs — caching, model routing, compression. Up to 40% savings, zero code changes. 0 users. 0 revenue. Just shipped the MVP. Gonna share the real numbers every week here.
Browser-based PDF tools, card maker, JSON formatter, and image utilities that work offline without uploading files.
@jainarun889 · X
Our Wikawe universe products. Check it out now... - loading
Access multiple AI agents through a unified API with per-use billing.
@agentsky_dev · X
Built for agents, not just models. AgentSky is live: one API to run Claude Code, Codex, DeepSeek, Kimi, OpenCode, and more in the cloud. Connect your real tools, then compare time, cost, and tokens side by side in Agent Playground. Try it:
AI-powered workflow automation platform to connect apps and automate tasks without coding.
ankit100 · HN
viaSocket – AI-built workflows, and your apps as an MCP server
Access Reddit data via API with free credits for building AI agents.
@thread_snoop · X
Give you claude/codex free access to search through reddit :)