Better Stack
@BetterStackHQ
Tracing reimagined with eBPF and OpenTelemetry.
Instrument your clusters without any code changes.
Get an AI SRE and a robust MCP server.
100 Following 11.1K Followers
Excellent deep dive into Needle from Andris and the @BetterStackHQ team!
Want to know what is new with Solid 2.0?
@JollyShopland with @BetterStackHQ has got you covered:
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.
Show more
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.
Show more
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.
Thanks @BetterStackHQ ^^
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
Show more
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
Show more
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 💪
Show more
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.
Show more
Stacked PRs are now built into GitHub: your 40-file monster becomes a chain of small PRs that review and merge independently.
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.
Show more
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.
Show more
Postgres 19 runs graph queries natively, so the five table joins you write to answer one question is now optional.
html-in-canvas is basically shaders for HTML, and Canvas UI is and awesome library that makes it easy to actually try.
@BetterStackHQ has the right idea with their MCP server - OAuth and API routes depending on the use case.
this video about gitreverse from @BetterStackHQ is so on point and explained best
thanks for the shoutout
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.
TypeScript 7 is officially released, and the Go compiler is absurdly fast.
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.
Show more
