AI Product Launches
What is AICrier?
Live AI developer news, ranked and linked to original sources.
> ▌
LAUNCH×
FRIDAY // 2026-08-28
34 itemsAUG 28
PT25M28S
// WATCH
NVIDIA Should Be Scared
+4
Theo - t3․gg
PT24S
// WATCH
Praxist: a tool where each generation of attempts starts from what the last one learned
+1
Github Awesome
PT13M11S
// WATCH
OpenAI Just Confirmed AGI Is Coming This Year
+7
AI Revolution
PT3M18S
// WATCH
Create Slides, Docs, and Templates
OpenAI
PT1M55S
// WATCH
Use ChatGPT Work on Your Phone
OpenAI
PT1M51S
// WATCH
Talk to ChatGPT Work
+1
OpenAI
PT2M29S
// WATCH
Use Your Computer and Browser
+1
OpenAI
PT2M21S
// WATCH
Build a Shareable Site
+1
OpenAI
PT2M10S
// WATCH
Scheduled Tasks
OpenAI
PT2M46S
// WATCH
Plugins & Skills
OpenAI
PT3M54S
// WATCH
Getting Started with ChatGPT Work
OpenAI
PT2M17S
// WATCH
Rime Finally Made Voice Agents Good Enough
Cole Medin
PT7M23S
// WATCH
RIP Hermes Agent: NanoClaw Is Lighter, Safer And FREE
+2
Income stream surfers
PT36S
// WATCH
Ox Alpha Is INSANE
+2
Theo - t3․gg
PT8M13S
// WATCH
Generate playable HTML5 games with Mistral Medium and GLM
+2
Mistral AI
PT1M45S
// WATCH
GLM 5.3 Flash vs Qwen 3.8 Flash: Cost & Quality! #shorts
+2
Income stream surfers
PT32M38S
// WATCH
One Residual-Stream Vector Controls AI Tool Calls
+3
Discover AI
PT40S
// WATCH
AI coders make how much?
The PrimeTime
PT12M10S
// WATCH
Solving Git At Scale
The PrimeTime
PT7M7S
// WATCH
This Makes Claude Code Free Forever
+2
Better Stack
LAUNCH// 3h ago
PRAXIST launches generational autonomous R&D
PRAXIST is a source-available autonomous R&D system that coordinates parallel agents, evaluates competing implementations, and carries validated evidence—including failures—into future generations. It integrates with Codex and Claude Code for repeatable research workflows.
praxistagentcoding-agentevaluationtool-useautomationdevtool+7+6+5+4+3+2+1
LAUNCH// 6h ago
Orange AI Computer Enters Public Preview
Orange AI Computer’s public preview is a Windows-first, local-first system for persistent memory, multi-model routing, agent orchestration, MCP tools, and evidence-backed execution. It connects Codex, Claude Code, local models, and optional AI hardware while keeping operational state on the user’s machines.
orange-ai-computeropen-sourcelocal-firstagentmcpai-codingdevtool+7+6+5+4+3+2+1
LAUNCH// 8h ago
Auris Debuts Interactive Car Reveal Hero
Auris is a concept automotive landing-page demo built by Thareek Anvar M. in Bolt.new. Its cursor-controlled reveal sweeps between a technical blueprint and a photorealistic Aurora GT render.
aurisbolt-newai-codingcode-generationdevtoolno-code+6+5+4+3+2+1
LAUNCH// 13h ago
Claudeforce puts Salesforce sales work inside Claude
Salesforce and Anthropic launched Claudeforce, beginning with Salesforce in Claude, a plugin offering 37 sales skills for analyzing live revenue data, updating pipelines, and taking governed actions. It is available to pilot customers, with open beta planned for September 2026.
claudeforceagenttool-usemcpautomationcloud+6+5+4+3+2+1
LAUNCH// 17h ago
Talus v2.0 brings verifiable agent execution
Talus Protocol v2.0 is now live on Sui mainnet, adding standardized agent identities, capabilities, execution payments, onchain proof verification, failure handling, backward-compatible upgrades, and priority fees. The release positions Talus as infrastructure for autonomous agents handling real value.
talus-protocolagenttool-useguardrailsautomationsdk+6+5+4+3+2+1
LAUNCH// 15h ago
CTRL Micro launches haptic agent deck
CTRL Micro turns an iPhone or iPad into a tactile control deck for Mac, letting developers monitor Codex, Claude, and Cursor, control desktop apps, view their screen, and dictate locally with Whisper. It supports nearby and end-to-end encrypted remote connections, plus custom controls for other tools.
ctrl-microagentcoding-agentcomputer-usesttdevtoollocal-first+7+6+5+4+3+2+1
LAUNCH// 15h ago
Revalvo launches local-first prompt eval workbench
Revalvo is a local-first workbench for running prompts across multiple LLMs, scoring outputs with 40 evaluators, versioning prompt changes, and batch-testing datasets. It keeps API keys and data in the browser, targeting developers who want evaluation before production.
revalvoprompt-engineeringevaluationllmtestinglocal-firstdevtool+7+6+5+4+3+2+1
LAUNCH// 15h ago
Play with Putty Brings Multiplayer Vibe Coding
Google Labs launched Play with Putty, an experimental workspace where multiple people use natural-language prompts to build tools and websites together in real time. Access is currently limited through a waitlist.
play-with-puttyai-codingcode-generationno-codedevtoolhosted-service+6+5+4+3+2+1
LAUNCH// 14h ago
Microduck launches $399 open-source RL robot
Hugging Face and Pollen Robotics introduced Microduck, a 25cm bipedal robot built for reinforcement-learning experimentation. It includes seven trained behaviors, an Apache 2.0 software stack, and tools for sim-to-real training and deployment.
microduckroboticstrainingopen-sourceedge-airesearch+6+5+4+3+2+1
LAUNCH// 15h ago
Aramb launches infrastructure for monetized AI agents
Aramb combines a no-code console and typed SDK for building long-lived AI agents with streaming, per-user isolation, workflows, tool integrations, and usage metering. Its pitch is moving developers from prototype to customer-facing, billable agent quickly.
arambagentagent-memorytool-usesdkapiclouddevtool+8+7+6+5+4+3+2+1
LAUNCH// 15h ago
Glisio launches Mac recorder with auto-zoom
Glisio combines click-aware screen recording, system and microphone audio, optional webcam capture, and a built-in screenshot editor in a local-first Mac app. Free exports carry a watermark; Pro costs $79 lifetime or $9.99 monthly.
glisiodevtoollocal-firstautomation+4+3+2+1
LAUNCH// 16h ago
PageIndex Launches Auditable Document AI
PageIndex launches a professional AI co-reader for long reports, filings, research papers, and technical documents, with citations that jump to highlighted source lines. Its structure-aware retrieval connects answers across documents while preserving verifiable evidence.
pageindexraglong-contextsearchmcpcontext-engineering+6+5+4+3+2+1
LAUNCH// 15h ago
Almanac Turns Company Context Into Agent Action
Almanac is an always-on AI agent that connects to company tools, builds a self-updating, source-backed wiki, and executes tasks through Slack, iMessage, its own browser, and terminal. The YC-backed product aims to make persistent organizational context the foundation for practical automation.
almanacagentagent-memorycomputer-usetool-useragautomation+7+6+5+4+3+2+1
LAUNCH// 16h ago
Caddi Launches Agents From Screen Shares
Caddi launched an agent that discovers repetitive back-office work, learns workflows through narrated screenshares in Loop Studio, and builds governed agents across existing business applications. Its hybrid architecture uses AI for judgment, deterministic code for exact execution, with scoped permissions and replayable logs. [Announcement](https://www.trycaddi.com/blog/caddi-launches-agent-that-builds-back-office-agents)
caddiagentautomationno-codecomputer-usetool-useobservability+7+6+5+4+3+2+1