AI Developer Tools & SDKs
-
KugelAudio is a self-hostable text-to-speech model with 39ms time-to-first-audio
Voice agents in regulated industries face a problem most TTS vendors ignore: you cannot ship customer data to a third-party cloud. KugelAudio, built by a four-person Berlin team and accepted into Y Combinator’s Spring 2026 batch, is a real-time text-to-speech model designed to run on your own infrastructure. ## Low latency, self-hosted KugelAudio reports a… Continue reading
-
Minimi feeds Claude live context from everything you do on your Mac
Most AI assistants only know what you paste into them. Minimi, a newly launched Mac app, takes the opposite approach: it quietly watches what you do across your computer and feeds that to Claude as live context, so you stop re-explaining yourself every session. ## How Minimi works Minimi captures the things that make up… Continue reading
-
Cresta Conductor builds production AI agents 2x faster from real customer conversations
Building a production-grade conversational AI agent usually means weeks of prompt writing, orchestration, and testing before anything is safe to ship. Cresta Conductor, launched June 11, 2026, attacks that by turning agent development itself into an agent — a “building-agent” that designs other agents on the Cresta platform. ## What Cresta Conductor does Conductor starts… Continue reading
-
Browse.sh gives AI agents muscle memory so they stop forgetting web workflows
Most AI agents wake up with browser amnesia. Every session they re-figure-out how to log in, navigate a dashboard, or fill a form, then forget it all by the next run. Browse.sh attacks that problem directly by giving agents muscle memory for the web. ## How Browse.sh works It is an open catalog of browser… Continue reading
-
Respan unifies LLM observability, evals, and gateway for 100+ AI startups
Teams shipping LLM features usually bolt together three or four separate tools: one for logging, another for evaluations, a third to route between providers. Respan, a Y Combinator-backed platform formerly known as Keywords AI, folds all of that into a single LLM engineering layer for observability, evals, prompt optimization, and an AI gateway. ## What… Continue reading
-
Tabstack Gives AI Agents Structured Web Data Without a Scraper
Getting clean data out of a webpage is still where a lot of agent pipelines break — you write a scraper, a site changes, and it quietly stops working. Tabstack, built at Mozilla, replaces that with an API: pass a URL and a schema, and get back JSON that matches every time, no scraper to… Continue reading
-
FluidDocs Deck Builder Turns a One-Line Prompt Into a Real Editable HTML Deck
Ask most AI tools for a slide deck and you get a document chopped into slides — right words, wrong structure. FluidDocs Deck Builder is the maker’s answer to that: describe a deck in a sentence and get back a real, editable HTML deck for a pitch, sales call, launch, or all-hands. ## What FluidDocs… Continue reading
-
Spotlight by Backplanes Grades What Your Coding Agent Actually Did
Coding agents like Claude Code and Codex do a lot in a single session, and most of it scrolls past too fast to learn from. Spotlight, from Backplanes, reads those sessions and turns them into reports — what the agent actually did, what’s worth keeping, what to fix, and where you could save time next… Continue reading
-
Publora Is a Publishing API That Lets AI Agents Post Everywhere at Once
If you want an AI agent to actually publish — not just draft — social posts, you hit a wall of a dozen different platform APIs. Publora, a publishing API built for the agent era, collapses that into one call. A single HTTPS request posts to twelve social platforms, and an MCP server lets agents… Continue reading
-
AgentOS Lets You Run a Fleet of AI Agents Like a Company
Running one AI agent is easy; running many across real projects gets messy fast — which agent did what, on which task, with whose approval. AgentOS, an open-source project from SapienX, is a control surface for exactly that. It’s built on OpenClaw, which handles the agent runtime, and adds the human operating layer on top.… Continue reading
