Give AI agents access to the live web with one API. Search & Fetch are free. Start at: https://teapaw.org/t.co/jGnj2ampDQ
Palo Alto, CA
Joined February 2024
- Tweets2.8K
- Following75
- Followers13.2K
- Likes4.5K
Pinned Tweet
The research behind every big decision usually takes a week. Claim that week back.
Today, we are introducing TinyResearch and TinySet.
Available now: tinyfish.ai/deep-work
TinyResearch turns a hard question into a report you can make a decision on.
TinySet turns a natural language prompt into a complete dataset, built from information scattered across pages on the web.
TinyFish retweeted
Happy Diwali to all our friends🪔❤️
@Tiny_Fish @adri_guha put on a stunning Diwali party last night 🙌
Thank you to everyone who joined us!
TinyFish retweeted
We built TinyResearch mostly by taking things away from the agent.
Ranking + extraction → plain code
Raw pages → out of the agent’s context
Separate planner → gone (it lost details on follow-ups)
Watching researchers use it changed what “good” meant.
tinyfish.ai/blog/inside-tiny…
TinyFish retweeted
last night at off the clock reminded me why i do this work.
when a room is curated with care, the right people find each other. the conversations go further. you leave with names you actually want to follow up with.
thank you to everyone who showed up, and to @Tiny_Fish @merge_api @vorfluxai Fuse and @MongoDB for building that room with us at sōhn.
saturday we keep going:
- badminton with @mireyehq and @vorfluxai
- diwali by the bay
more than happy to have you at both. links below. – at San Francisco, CA
TinyFish retweeted
Our community team is growing!
@adri_guha has joined me to scale our community function as TinyFish grows.
This was not an easy hire. Running community as a team of two is like a group project where you present four times a week, every week. The role demands extreme attention to detail, adaptability, ownership, and being fluent in social settings.
Community is a highly visible function. To succeed, it takes two people who think alike, take ownership of every detail, and care deeply about the tech ecosystem.
We've already accomplished a lot in a short time, and I have no doubt there's much more ahead for us together.
So excited to have you by my side, Adrika!
TinyFish retweeted
Founders in NYC:
@Tiny_Fish is hosting a private dinner with @gen_legal_inc and @hydra_db
We will be inviting our friends and also welcoming new faces to join us.
Register below to meet our team in person in NYC for our first private event!
TinyFish retweeted
We're hiring Ecosystem Engineers at @tiny_fish!
If you love talking to people, have a technical background, and aspire to be a founder, or have already been one, I’d love to meet you.
This is an internship/part time role to start. You’ll work directly with me to help AI native startups integrate TinyFish into their products and build lasting relationships with founders and engineers across the ecosystem.
What you’ll work on:
- TinyFish integrations with AI native startups.
- Work alongside their teams to build and launch those integrations.
- Develop partnerships and support the teams beyond the product
You’d be a great fit if you:
- Have a CS background and enjoy building.
- Are plugged into the tech ecosystem through a university club, VC network, AI startup, or builder community.
- Take work personally
- Based in SF or nearby and available to start immediately.
We host 3–4 events every week, from founder and researcher dinners to builder gatherings. This role gives you direct access to our ecosystem, events, and partners.
If this sounds like something you’d be interested in, send me a message.
Want to meet our community? Join us at an upcoming event in SF and NYC links below.
Ft. the co-host team from our Jazz cafe event last night!
TinySet goes wide. Describe the dataset in one sentence and edit the columns it plans. It splits the job into smaller pieces, like every city or category, and fills each row only from pages it actually reads.
2
26
2,516
TinyResearch and TinySet are available via the ChatGPT Plugin, Claude Connector, TinyFish Playground, API, MCP, CLI, and SDKs.
Documentation: docs.tinyfish.ai
Learn more: tinyfish.ai/blog/introducing…
25
1,672
TinyFish retweeted
Best community building you can do is build a fkin good product
I use @Tiny_Fish everyday so when i saw @DeonMen is hosting an event, it was a no brainer
- best curated room
- great food
- free credits to a tool my claude actually loves
Ps: we ripped pokemon cards too (thanks to @0xAlpaca_)
Sci-fi promised glasses that tell you what you’re looking at.
The hard part was never “what is it.” It’s “what’s true right now.”
NewEyes by @CollovLabs closes that gap with TinyFish Search + Fetch.
Full Story👇
tinyfish.ai/blog/collov-newe…
TinyFish retweeted
Me and @Tiny_Fish Hosted a Dinner with Creators (5 M+ Followers) and Founders at Trident Hotel Near Devcon 8
More than 63 People arrived in a venue of 50 people😭 Meaningful conversations for about 4 hours.
We had talks from 8+ people and it was all about how to grow your content engine and company!
“This has to be the best and most meaningful Devcon 8 Side event” - Christian
We are hosting more events soon!!!
When we recorded this with @Saboo_Shubham_ , we were at 400K users.
We’re well past half a million users now, and our co-founder and CPO @shuhao_friday spills exactly how we got there: youtube.com/watch?v=vCHe1CSE…
A lot of our users aren’t people. They’re agents, and they’re the pickiest customers we’ve ever had.
So we built for them:
→ Killed subscriptions, because no agent wants a $20/month plan for a 40 cent job
→ Trained our own web-native model and cut costs about 10x
→ Put our pricing and benchmarks right where agents look before they pick a tool
→ Let agents sign up without bugging their human
Our favorite moment is Shuhao looking at the camera and telling any agent listening that it doesn’t need its human to open a TinyFish account.
Full episode is worth the watch, for humans and agents alike.
TinyFish retweeted
New on @tenki over the past couple of weeks:
4 new pre-built sandbox images, ready to run:
- @ampcode: a coding agent
- @PrimeIntellect's Prime Agent: a headless agent with a persistent Python kernel
- @Tiny_Fish: web agents, remote browsers, fetch and search
- @hotdata_dev: an instant SQL database for each agent task
Control what a sandbox can reach:
- Named egress policies, saved once per workspace
- HTTP rules by domain, method and path
- HTTPS inspection
- A new Egress page in the dashboard
Code Reviewer now works on @cursor_ai Origin, as well as GitHub. Connect from Tenki or from Cursor's App Marketplace, and reviews show up as inline threads with a verdict.
Preview URLs now belong to your workspace, not to one sandbox. Unbind a URL before you terminate the session, bind it to the next one, and the hostname stays the same. That's useful for OAuth callbacks and webhooks.
Also:
- Template builds show what each build uses, and you can delete builds one at a time
- Every Add Funds top-up now comes with a Stripe invoice
- Sticky disks: $0.20/GiB-mo
tenki.cloud
DeepSeek, Kimi, GLM and every other open model on @FireworksAI_HQ can now search the live web with TinyFish, for free.
TinyFish is officially a web search partner for the Fireworks ecosystem.
The setup is a single tool definition: docs.tinyfish.ai/integration…
Fresh results from the web, reasoned over at Fireworks speed.
TinyFish retweeted
“The coding interview is dead!” We have turned the idea by @felixrieseberg at @AnthropicAI into a huge AI-Native Talent Hunt at SF Tech Week.
Don’t interview people about the work. Watch them do the work.
We believe hiring needs a better signal than LeetCode: Proof of Work with your PROCESS, a new credential in the AI era.
We made it happen with early believers in changing how AI-native talent is assessed and hired. Thank you to our amazing co-host @Tiny_Fish, agent partner @OpenAI Codex, and hiring partners @GenDigitalInc, @ZingageAI, @Gruve_ai, @cresta, @CreaoAI, @SmithfieldFoods, @dify_ai, @OpusClip, and @openart_ai.
Together with the enthusiasm of our candidates, we got:
• 1,600+ signups
• 300+ attendees
• 200+ submissions
• 10,000+ minutes working videos
How it works:
1. Companies turn internal business problems into real work sample tasks on NeoWork (neo-work.ai).
2. Candidates tackle them on NeoHuman (neo-human.ai) using Codex and other AI tools.
3. NeoWork evaluates the deliverable and recorded process. We highlight moments that reveal human judgment: catching an AI mistake, checking a claim before acting, or changing an approach that isn’t working.
4. Hiring managers review the work, our assessment, and supporting evidence to decide whom to interview or hire.
We grouped the 10 companies' works into 3 tracks: Product, Marketing, and Engineering. Each work was designed for about an hour. Examples include:
@GenDigitalInc: design a browser agent’s action pipeline that resists prompt injection.
@CreaoAI: turn design guidelines into enforceable rules and prove they catch real UI issues.
@cresta: build a support agent that follows policy and knows when to escalate.
@openart_ai: select consistent memories for an AI director.
@dify_ai: decide whether a support AI is ready to ship, backed by evidence.
Congratulations to our overall winners!
🥇 @samshanmukh: 97/100
🥈 Rohan R: 89/100
🥉 Shivam Mahajan: 87/100
And our track winners:
Product: Marcus Wong
Marketing: @AndrewDrok
Engineering: Akshat Patel
We know today’s job market is tough, even for people already in work:
- Juniors need experience to get experience.
- Career changers have skills their old titles don’t show.
- Laid-off professionals have to prove themselves all over again.
- Employed professionals must adapt to AI while keeping up with today’s workload.
That’s why we’ll keep hosting AI-Native Talent Hunts, with campus events at @UCBerkeley and @Stanford this November.
Follow our events calendar (lnkd.in/gzjNWahB) for your next chance to show what you can do.
