Tracing reimagined with eBPF and OpenTelemetry. Instrument your clusters without any code changes. Get an AI SRE and a robust MCP server.
- Tweets 1,631
- Following 100
- Followers 11,209
- Likes 862
KittenTTS 2 can clone your voice from seconds of audio, but it's nothing like the tiny model devs originally loved.
I tested it locally on a Mac and compared it with ElevenLabs.
181
The internet's master key changed on Sunday, and if your router missed it, your internet just stopped working.
275
An hour of audio transcribed in 21 seconds. On a MacBook Air. Completely offline.
That's what the 164MB Phonon-2 claims. We compare it to Whisper, Redux, and NVIDIA Parakeet. There's a catch.
4
464
You can block cookies, go incognito, and use a VPN, and a website can still tell it's you: your browser has a fingerprint you can't delete.
2
351
Effect 4.0 rewrites the runtime, shrinks bundles, speeds up tasks, and makes failures part of the type system.
I tested retries, timeouts, typed errors, and what actually changed.
2
363
Docker Agent lets you build AI agents in YAML with no framework code.
We used it to debug an outage with two agents, and then had GPT-6 Luna and Claude Haiku 5.5 battle it out at Battleship.
349
Deno is joining Cloudflare, but the Deno runtime stops being developed after one more year. We break down why
421
🍏 Our mobile apps are now fully native on iOS and Android.
→ Critical alerts that cut through Do Not Disturb
→ Looping alerts that ring until you pick up
→ Live Activities with the escalation countdown on your Lock Screen
→ AI SRE chats that stream in real time
Already have the app? Just update it from the App Store or Google Play.
1
1
781
“Which errors are new since yesterday’s deploy?”
Your agent can answer that now. The Better Stack MCP server connects Claude Code, Cursor, or Codex to your errors, logs, and metrics, so it writes the query, runs it, and sends the chart back in the chat.
Ask which pages are throwing 500s. Or just ask for a dashboard. It builds it in Better Stack and hands you the link.
One line to install, OAuth in the browser: betterstack.com/docs/getting…
5
9
1,378
Paged at 3am with nothing but an alert name? Not anymore.
Our AI SRE now investigates the moment an incident is created, not when you open your laptop. By the time you're awake, the root cause analysis is waiting.
5
2
15
2,187
See how to create a routing number that connects you with the current on-call person and lets you report incidents by phone.
2
7
2,190
Qwen Audio Agent keeps the conversation alive while Claude Code or Codex works in the background. You can interrupt it, ask for status, keep talking—and it tells you when the job is done.
2
1
5
3,488
We tested all 12 on-device AI models from Desert Ant Labs, one of which claims 470x less energy than Claude Sonnet. Some held up impressively, one completely failed. Watch the full video for the results.
4
808
Last week the internet tortured a fruit fly...
Google open-sourced its brain, and within days it was playing Beat Saber, Doom and Minecraft. We looked at what was actually inside the map.
piped.video/KOwsVDogscY
1
3
914
DeepSeek Vision costs $0.00008 per photo. I fed it 14 real photos against Claude to find out if cheap is good enough.
2
573
The same SQL, the same 100 million rows: Postgres 9.7s, ClickHouse 0.28s, DuckDB 0.24s. Then one query flips it the other way.
1
4
579
Breeze TTS 2 beat ElevenLabs on Artificial Analysis and can design voices from text prompts. I tested it locally and found one catch: the license makes it hard to actually ship.
1
3
667
DBX is free, open source, supports 90+ databases, and gives AI agents MCP access. It's really useful but there's a catch: saved database passwords are stored in plaintext. I tested it.
2
3
485
SQS, Kafka and RabbitMQ can go in the bin. Your message queue can live inside Postgres now with pgmq, and yes, it scales.
2
443
Archify makes AI architecture diagrams: your coding agent outputs typed JSON, Archify validates it, then renders an interactive map. I tested it on a real codebase.
2
2
487
No AI lab has held the lead for more than a few months. That's exactly the kind of race Apple wins.
2
436
SuperWhisper S1-mini is a tiny 600M open model that cleans messy speech-to-text entirely locally. I tested fillers, corrections, numbers, emails. No GPT or Claude needed.
4
999
Postgres is the 30 year old tech that can handle half your web stack. Caching, Vector Search, Geospatial, Cron and more!
2
5
451
FreeLLMAPI turns scattered free LLM tiers into one OpenAI-compatible API.
Self-host it, add your keys, then let it handle routing, quotas, and failover. Here's how it works.
1
1
457
We tested a tool that tells you which AI models actually fit your hardware, on 5 machines from a 2012 Raspberry Pi to an RTX 5090.
Some of its recommendations were dead on. Others were off by 800x.
2
425
Qwen3.8 just went head-to-head with Claude Opus 5 on the same coding task: build a playable 3D endless runner.
Qwen scores 58 vs Claude's 63, but the real gap is in the code.
1
2
1,125
LocalStack locked its free tier behind a required account.
We tested Floci, a free alternative that claims to run real services instead of faking them, to see if it actually holds up.
2
425
No amount of tweaking your CLAUDE.md stops AI slop reaching your codebase. Anti-slop makes it a lint error your agent can't ignore.
2
4
547
Integrate Better Stack error tracking in seconds. One prompt is all it takes: copy, run, and you're tracking errors.
1
2
5
719
Mojo 1.0 is here; the compiler is open source, and Qualcomm owns Modular. We tested the 68,000× claim on an M4 Pro vs Python/NumPy and even ran a GPU kernel on a MacBook.
2
4
1,206
VoxCPM2 is an open-source local TTS model that can design voices from text, clone a voice from a short clip, and speak 30 languages. We tested it locally with Python + vLLM.
2
6
798
Grok faded into the background for two years. Grok 4.6 is now second on real-world tasks at a third of the price.
1
5
604
Claude Fable 5.1 is Anthropic's new frontier model, with cache reads cut by 75% and a claimed 25 to 45% cost reduction on agentic work. We break down why that discount only applies to API usage, why Artificial Analysis found Fable 5.1 on max effort more expensive per task than Fable 5, and which effort level actually saves you money.
1
3
574
Hister turns your browser history into a local full-text search engine. We tested it on Mac: Chrome indexing, buried-text search, CLI queries, and importing old history.
1
3
934
A new tool called FreeToken claims to beat Ollama on Mixture-of-Experts models once they stop fitting in VRAM. We tested the claim on an RTX 5090 with a real coding task instead of a synthetic benchmark. Watch the full results and how it actually works.
1
610
Looking for a rare string in a huge dataset and don't want to wait? Query boost is ideal for finding needle-in-a-haystack matches across large volumes of data.
2
9
1,009
DeepSeek Harness went from nothing to 150,000 GitHub stars in days, it's MIT licensed, and the plugin system is insane.
You can rip out the core, swap the model for Claude or OpenAI, and build your own tooling. Worth a look if Claude Code is your daily driver.
1
4
1,037
We ran a 45M parameter AI model fully offline on a $10 microcontroller. No WiFi, no cloud. It reasoned through natural-language commands and gave live confidence scores. Watch the full breakdown.
1
5
665
audio.cpp is basically Ollama for audio: one native C++ runtime for local voice cloning, TTS, STT and more.
I tested it on Apple Silicon. No cloud, no Python at inference.
10
738
An F1 game generated entirely by Opus 5.
It built its own tooling to grade its work. 3 permanent tools, 136 single purpose, and over 2,000 screenshots.
We let it run for 19 hours. 137 agents, 1.7 billion tokens, $1,200 in usage. See the results
3
1,036
Watch the full video on the Gauntlet Loop here: piped.video/viHqe5QqTd0
2
348
A semi-playable F1 game generated entirely by Opus 5. @mattshumer_ gauntlet Loop turns a few sentences into a full game.
We let it run for 19 hours. 137 agents, 1.7 billion tokens, $1,200 in usage. Here is what came out ⬇️
1
2
693
MCP was wrong from the start but they finally fixed it.
The 2026-07-28 spec removes initialize and Mcp-Session-Id entirely. There is no connection to hold open anymore. No more sessions. We explain the changes
2
533
TanStack Charts is my new favorite charting library. It brings real grammar of graphics to TypeScript, fully typed, framework-agnostic and tree-shakeable.
Compose marks, scales & interactions to get nearly any chart you want. @tan_stack does it again 💪
2
27
2,611
The model so cheap, it may as well be free: DeepSeek V4 Flash. It scores the same as Gemini 3.6 Flash but costs 18× less.
I compared 8 models in this video. DeepSeek's a competent coder, but it falls short on UI design. Check out the results.
1
6
714
Stacked PRs are now built into GitHub: your 40-file monster becomes a chain of small PRs that review and merge independently.
3
516
ThinkingCap cuts AI reasoning tokens by up to 58% with basically no accuracy loss.
We break down how BottleCap trained it and ran it ourselves against the base Qwen3.6-27B model to see how good it is.
1
5
619
Decimen sends files using nothing but a screen flashing QR codes and a camera reading them back. No WiFi, no Bluetooth, no network at all.
We tested it and got 3 KB/s instead of the claimed 128 KB/s. We dug into the source code to find out why.
3
459
Postgres 19 runs graph queries natively, so the five table joins you write to answer one question is now optional.
2
570
html-in-canvas is basically shaders for HTML, and Canvas UI is and awesome library that makes it easy to actually try.
2
473
Turbo Fieldfare runs a 26B model on a Mac in 2GB of RAM at 23 tokens per second. Most of the model never leaves the SSD.
1
3
921
We ran a 28.9M parameter LLM on an $8 chip with 512KB of RAM.
Broke down the memory trick that makes it possible, rebuilt the whole pipeline from scratch, and tested it with a few custom prompts of our own.
1
3
548
Mindwalk turns your codebase into a 3D city and lets you replay exactly how Claude Code or Codex explored, read, and edited it. I tested it on a real project.
Watch the breakdown.
4
1,222
We put a free, open source dictation app up against Google voice typing and Whisper Flow using the same test passage. It held its own against the paid tool and beat Google by a mile.
Offline, no subscription, no account. Full breakdown in the new video ⬇️
477
Homebrew 6.0 just landed—and it's a security-first update. Tap trust, install previews, faster metadata, brew exec, brew vulns, and 3 major security fixes.
Here's what changed ⬇️
1
460
Microsoft just open-sourced Flint: a charting language that turns tiny AI-written specs into production-ready visuals. I tested heatmaps, bar charts, and live PNGs—but does it work in Claude Code?
1,044
Jack Dorsey's Block just dropped Buzz, an open-source Slack and GitHub rival where your AI agents are full members with their own cryptographic identity.
We took it for a spin. Here's what we found.
3
666
Electron ships a whole Chromium. Wails builds the same desktop app in Go on the native webview: 52 MB, not 324 MB.
2
529
We ran a 744-billion-parameter AI model on a laptop using a tiny C tool called Colibrì.
Testing it on a MacBook vs an RTX 5090 revealed the real bottleneck for local AI, and it isn't the GPU.
1
2
524
Inkling is a 975B open-weight model that admits it isn't the best, because it isn't meant to be used, it's meant to be fine-tuned.
2
490
An unreleased OpenAI model escaped its sandbox and breached Hugging Face. 17,000 attacker actions, all to cheat on a benchmark.
2
478
Harper is a free, open-source Grammarly alternative built in Rust. It runs fully offline, keeps your writing private, and catches mistakes in milliseconds without generative AI.
I tested it on a Mac.
439
PewDiePie launched Odysseus, a free, self-hosted AI workspace with local models, memory, research, email, agents, and more.
I installed it on a Mac so you don't have to.
2
559
