AI agents you can delegate real work to. Connect your tools, automate work 24x7, and share agents across your whole company.
San Francisco, CA
- Tweets 338
- Following 0
- Followers 3,167
- Likes 435
Pinned Tweet
Lee, Founder and Owner of @FitFoodieFinds, scaled Meta ads from 3 to 60 with Tasklet agents.
Her team of four women in Minneapolis creates the recipes, photos, and videos. Agents handle the work behind the scenes, giving the team more time for the creative work.
Full story ↓
3
5
11
2,585
GPT-6.1 Sol from OpenAI is live in Tasklet! ☀️
Your knowledge, connections, automations, and team can stay put – choose from all the best models to power your agents.
Better cost efficiency and increased intelligence make 6.1 Sol great for your long-running agents and threads!
2
6
1,121
One platform, every major model.
Feel free to give Sonnet 5.5 a try in any of your agents without rebuilding anything!
2
1
5
584
Tasklet retweeted
RIP tableau
we're testing new onboardings @taskletai - i had one prompt: "figure out conversion across onboarding cohorts"
tasklet used our team shared connecters to dig into code, rollouts, ads/analytics and build a custom dashboard
2
2
7
401
Tasklet retweeted
everyone is claiming cost is for vms, but it really is inference - if you architect it right.
there is a trend to run every agent in it's own vm - but that locks you into an inflexible and expensive architecture
at @taskletai we flipped it: every agent has access to a vm, but we spin them up on demand for each session, backed by a virtual, infinitely scalable file system. keeps cost down and allows us to do things like shared drives and cross-thread file access.
i'll post more details on the file system soon
Hearing this week that personal assistants cost $3k-$7k per user per year, depending on scale effects / vertical integration / model selection. Implications/questions:
- Consumer subsidies will be a huge factor; people don’t care about privacy, they want utility and they want it for free.
- Model costs are deflating, but not fast enough for many startups. The dark horse is that Jev may rapidly deflate computer-use costs 100x.
- An interesting counter-position would be building an expensive and narrowly focused agent — perhaps combining proprietary supply (Carbone tables!) with a unique POV (“no one goes there anymore, go to The Grill instead”).
- Social is wide open but hinges on a key design challenge: how can agents increase social cohesion and not feel invasive in a group chat? A working social agent would rightly attract an avalanche of capital and be hard for incumbents to replicate.
- Finally, there are vertical focus areas like financial services where the LTVs outweigh the subsidy needed and are outside the near-term competitive focus of the incumbents. If it were me, I’d focus here — Credit Karma as an agent >> Credit Karma as an app.
Overall, it’s still very, very early, and we may very well be in the MySpace era of social or iPhone 2010. Hang onto your hats and keep building!!
2
7
332
Tasklet retweeted
one of the hidden features of @taskletai are ad-hoc dashboards. hook it up to any tool and create customized, interactive dashboards with a single prompt.
give it a try with gmail: "connect to my gmail and give me an interactive dashboard for reply stats, include filters by domain, sender and subject".
We're seeing a similar trend at @TaskletAI
Astra has been really popular. Also Gemini 3.8, Grok 4.6, and GLM 5.3 growing share recently.
2
1
6
390
OpenAI's GPT-6 Sol and GPT-6 Luna are both available in Tasklet ⚡️
Both bring advances from GPT-6 Astra, and cost 50% less to run in Tasklet than their GPT-5.6 versions.
2
6
496
Lee, Founder and Owner of @FitFoodieFinds, scaled Meta ads from 3 to 60 with Tasklet agents.
Her team of four women in Minneapolis creates the recipes, photos, and videos. Agents handle the work behind the scenes, giving the team more time for the creative work.
Full story ↓
3
5
11
2,585
The results:
✓ 3 → 60 Meta ads
✓ $10K–$15K/month saved by bringing work in-house
✓ Every ad checked every 15 minutes
✓ SEO research and approved updates
The agents “free up our creative brain space to really think forward and be excited for what's next.” – Emily
1
70
“Tasklet has allowed us to have that time back and to really focus on the best part of our job, which is creating content.”
– Lee Funke, Founder and Owner of @FitFoodieFinds
Read the full story and meet their agents: tasklet.ai/learn/case-studie…
1
57
Tasklet retweeted
Calling Houstonians who want to delegate to AI: I’ve been spending a lot of time learning how to work with AI agents. I’m still learning, and I'm sharing my learnings as I go.
A month ago, I trained 25 people across two sessions of "How To Build Your First AI Agent."
This Friday I’m hosting another small, hands-on workshop at @theion. Bring your laptop and we’ll actually build an agent using @TaskletAI, connect it to tools you use, and give it real work to do.
It’s free, but I’m keeping the group small and asking people to apply this time.
Friday, Sept. 11 • 2 PM • The Ion, Houston
Apply here: forms.gle/7Bdas9uGX1g9yCrLA
3
4
16
799
Tasklet retweeted
Tasklet’s multi-model harness is incredible for building rich, complex apps — especially games!
In the 90’s there was a game by Ambrosia Software called Escape Velocity that I was obsessed with. So I decided to recreate something similar with Tasklet.
In the game, you can fly around the galaxy warping from place to place. You can land on planets, fight pirates, take on secret missions, trade commodities and so much more. It’s a rich immersive world.
To build this in Tasklet, I used:
- Astra for most of the coding
- Fable 5.1 for most of the creative work — storylines, missions, characters, etc.
- OpenAI image models for the sprites, planets, avatars, etc.
- Gemini omni for video generation for cutscenes.
- GLM 5.3 to actually power the interactions on the game (for lower cost)
Tasklet’s instant apps feature made it effortless to actually build & serve the app, and Tasklet’s browser use allowed the agent to visually inspect it’s work to make sure everything looked right. For example, it spent a long time reviewing the landing sequence animations to make sure they were fun.
7
3
41
10,812
Does it actually matter which AI model you use? If one works, why switch?
GPT-6 Astra, Gemini 3.8 Flash, GLM 5.3 Flash, and Grok 4.6 are now in Tasklet.
3 reasons to try them: different strengths, better results, or lower costs.
1
10
4,599
3. Lower costs
Sorting emails or extracting invoice details? Try GLM Flash, a lower-cost option for simple repetitive tasks.
Tasklet’s plans are usage-based, not per-seat, and small savings on recurring work can add up across your team.
1
2
101
Another benefit of one platform with all the models: ChatGPT Images 2.5 came out today & it's already in Tasklet. Your agent can use it regardless of which model its on.
So if your current model works well, keep it. Not sure which to use? Ask your agent.
ChatGPT Images 2.5—faster, sharper, smarter, with better tools for creating whatever you can dream of.
- Faster image generation to keep your ideas flowing
- Improved fidelity for more natural, recognizable images
- Consistent details across multiple edits
- Comment-based edits to change only what you want
2
6
275