Joined March 2022
- Tweets 15,351
- Following 263
- Followers 60,126
- Likes 19,626
4,112 Photos and videos
Pinned Tweet
Jun 10
Most AI agents have expensive amnesia.
You run a task → it finishes → memory wiped.
Run it again → you pay full price.
You’re not paying for results.
You’re paying to re-teach the same task.
12
76
217
51,124
One of the biggest bottlenecks in AI video today isn't quality anymore. It's cost.
I've been hearing about an upcoming Dreamina release called Dreamina Seedance 2.0 mini, reportedly arriving on June 15, that could bring Seedance-level video generation to a much lower price point.
If the quality is anywhere close to Seedance 2.0, this could make high-end AI video creation accessible to a much larger group of creators.
Lower costs mean more experiments, more creativity, and more people building.
Definitely one to watch.
Give it a try 🔗 aisecret.co/dreaminaseedance…
#dreamina #dreaminaseedance2mini
2
32
34
17,524
codemarch retweeted
Jun 11
The absolute biggest bottleneck in the AI video landscape right now is no longer the output quality.
It all comes down to the price tag. Keep your eyes on this upcoming solution.
👉 aisecret.co/dreaminaseedance…
The timeline is buzzing about an upcoming Dreamina rollout that promises to bring premium, Seedance-tier video generation to a radically more accessible price point.
If the visual fidelity is anywhere close to the power of Seedance 2.0, this drop is going to unlock high-end AI cinematic creation for an entirely new wave of digital artists.
Slicing that financial barrier translates into one massive shift: limitless experimentation, unfiltered creativity, and way more builders bringing their visions to life. 🚀
Unquestionably a release you need to monitor closely.
#dreamina #dreaminaseedance2mini
2
13
18
10,513
Jun 10
TTS is fun to demo~ until you have real users.
That’s when the cost hits.
Inworld cutting TTS prices 50% (and giving double credits in June) actually makes it usable at scale.
Jun 10
We want to make AI accessible for everyone, so we're reducing our API prices by ~50%.
Consumer AI growth is still blocked by model costs. 97% of consumer AI users will never pay. But every session still consumes tokens.
This is the math we've found every consumer AI developer is stuck with. The model pricing they're forced to use, was built for $300/mo enterprise seats, not $10/month consumer subscriptions.
So we're taking a step to solve this with reduced pricing across TTS, STT and LLMs.
Teams like Wishroll, Biblechat, Talkpal, and Luvu are already seeing 40-95% reductions in their AI costs.
Full breakdown on our blog: tinyurl.com/InworldBlog
Pricing details in bio link.
1
31
63
15,649
Jun 10
Most AI agents have expensive amnesia.
You run a task → it finishes → memory wiped.
Run it again → you pay full price.
You’re not paying for results.
You’re paying to re-teach the same task.
12
76
217
51,124
Jun 10
Now compare that with the latest external benchmark:
AllyHub:
→ 13 credits
→ 500 reviews
Manus:
→ 241 credits
→ 60 reviews
~8x more reviews
at ~1/10 the cost.
That’s not just better output.
That’s a different cost curve.
1
7
1,554
Jun 10
The takeaway:
Most AI tools charge you for rediscovery.
AllyHub compounds efficiency instead.
200 free credits/day → try it: allyhub.ai
4
1,516
codemarch retweeted
Jun 9
that AI bill you flinch at every month? $100 ?
everyone’s been venting about it for months. so we spent 12 cents asking Ally what people are actually saying.
here’s what it did, untouched:
→ 15 search angles
→ 580 posts torn through on X reddit
→ every comment ranked and read
→ 500 steps, non-stop
→ full report, start to finish
the bill: 12 cents.
AllyHub — the first AI agent too cheap to bother counting is live.
one sentence in, full report out. next time it’s one click.
free credits drop daily — enough for 10–50 runs.
Grab your Ally — let it handle the tedious work you don’t want to touch.
105
154
607
1,028,629
codemarch retweeted
Jun 9
Analytics has been waiting for this kind of shift.
@MoraHQ makes it feel natural to ask complex data questions and actually trust the answers, with SQL you can verify.
Love how it removes the friction between questions, insights, and dashboards.
This is a strong step forward for AI-native analytics → mora.com
Jun 9
After chat and code, analytics is the biggest opportunity in AI and it's still completely untapped with our current tools.
As founders, data scientists and operators we deserve something better.
That's why today we're launching @morahq 👇🧵
9
32
85
20,756
Jun 9
This feels like a shift from creation to execution.
Creatify isn’t just generating ads anymore - it’s running campaigns.
Connecting creative, scheduling, and launch in one flow is a big deal.
That’s where most of the real work actually happens!
Jun 9
Creatify Agent, Wave 2. You watched it make the ad. Now watch it run the whole campaign. It learns your brand. Directs a cinematic spot — no slop. Connects to Meta, TikTok, and Google, and launches on the date you pick. The agent didn't get an update. It got promoted. Drop your product below.
31
59
10,493
codemarch retweeted
Jun 9
Creatify Agent, Wave 2. You watched it make the ad. Now watch it run the whole campaign. It learns your brand. Directs a cinematic spot — no slop. Connects to Meta, TikTok, and Google, and launches on the date you pick. The agent didn't get an update. It got promoted. Drop your product below.
198
395
746
2,057,775
Jun 9
Every time I ask Claude something complex, I get pages I barely skim.
Now with HyperFrames by HeyGen, that same answer becomes a video people will actually watch.
The shift isn’t the intelligence - it’s the format!
Hyperframes is now an official @claudeai connector
LLM answers are often dense pages of text that go unread
we partnered with anthropic to fix that, @HyperFrames_ runs as an mcp where you work
docs get skimmed. video gets understood
proof: this video was made with Hyperframes
3
31
62
16,544
Hyperframes is now an official @claudeai connector
LLM answers are often dense pages of text that go unread
we partnered with anthropic to fix that, @HyperFrames_ runs as an mcp where you work
docs get skimmed. video gets understood
proof: this video was made with Hyperframes
161
158
649
732,400
Jun 8
Really impressive launch.
Bringing deployment, compute, and models into one stack just makes sense.
#GMICloud #GMIAgentBox
Docs: docs.gmicloud.ai/agentbox-ma…
@gmi_cloud
Jun 8
Today, we are launching GMI Agent Box.
A complete infrastructure stack for production-ready AI agents:
native Docker, flexible deployment, 200 models under one API key, dedicated compute across regions, and a marketplace for distribution.
Available now.
42
6,364
codemarch retweeted
Jun 8
The future of AI is action, not just chat.
GMI Agent Box launches June 8.
Take a look at the video for a preview.
1
16
5,417
Jun 4
Really solid launch!
Consistent avatars dynamic cinematic scenes is a powerful combo
and api access makes it easy to plug into existing pipelines
Nice step forward for scalable video creation...
Cinematic_avatar api is live
keep your likeness, add cinematic range and build your video pipeline via your coding agent
install HeyGen CLI HyperFrames skill to create launch videos like ours
docs and cli setup in thread ↓
15
62
80
12,886
Jun 3
We’ve had a solid language for how brands look
not really for how they behave in motion
so when you ask for video, you get something that feels more like a layout than a sequence
frame.md feels like that missing layer!
Introducing frame.md, a spec built for videos & motion
design.md kept your brand consistent across screens
but when applied to videos, agents translated it back into webpages and decks
frame.md teaches your agents how to make branded video
turn your design.md into frame.md ↓
15
40
85
13,664