I love open source so much
I can just add @Tiny_Fish Search + Fetch to my own OpenCode setup, and now all my agents instantly get access to fast, free web search.
So cool. @thdxr@opencode
landed in the bay and my acting skills were immediately put to work. π€£π€£
thankfully it was for BigSet.
describe any dataset in plain english. get the data.
btw its open source too
GO CRAZYππ
What if you and your agent had all the data that always stays fresh?
Structured, on demand, never stale.
Introducing BigSet.
Describe the data you need in plain English β get a structured dataset built from the live web, that refreshes regularly.
It's live and open-source.
@jdsimcoe@Tiny_Fish@raycast hey jonathan, thanks for trying it out! make sure your script settings match the screenshot, and that youβve got the tinyfish cli installed + authenticated.
dm me if you run into any more issues, happy to help!
Why open a browser just to search the web?
Connect TinyFish's free Search to a @raycast script.
Now you can query the entire live internet right from your keyboard.
You've been asking for this one...
Now in preview: Codex in the ChatGPT mobile app.
Start new work, review outputs, steer execution, and approve next steps, all from the ChatGPT mobile app. Codex will keep running on your laptop, Mac mini, or devbox.
Your Hermes Agent or @openclaw can now write its own Skills and self-improve.
> Uses our free Search and Fetch to research any topic
> Writes ready-to-use SKILL.md
Send tinyskills to your agent https://t.co/UaTqNh6zt9
Grab your API key: https://t.co/FNHikQloq6
Introducing SubQ - a major breakthrough in LLM intelligence.
It is the first model built on a fully sub-quadratic sparse-attention architecture (SSA),
And the first frontier model with a 12 million token context window which is:
- 52x faster than FlashAttention at 1MM tokens
- Less than 5% the cost of Opus
Transformer-based LLMs waste compute by processing every possible relationship between words (standard attention).
Only a small fraction actually matter.
@subquadratic finds and focuses only on the ones that do.
That's nearly 1,000x less compute and a new way for LLMs to scale.
new this week in helium:
- web content frame with concentric rounded corners that dynamically adjust based on the adjacent surface across all platforms
- a whole lot of zen mode UX and stability improvements that make it a joy to use, including new options to pin the sidebar or the top bar. for example, you can keep the toolbar pinned, but reveal vertical tabs on hover!
- split view and side panel UX improvements that make all panels inside helium look consistent, coupled with dynamic rounded corners
- better default theming with brighter shades and cuter colors
- redesigned link hover bubble, which is now consistent across all platforms
- custom profile avatars (thanks to a community PR!)
- (almost) full translations to all languages thanks to the community efforts on github
- UI rendering fixes on windows that make helium look even better and consistent with other platforms
- security fixes (latest chromium)
- other small improvements that make helium better
you can get the latest release (0.11.5) on https://t.co/JEurq40fCr or github: https://t.co/OHj7pQjK5l