Anthropic should stop beating around the bush and commit to releasing Mythos, instead of building unnecessary hype around it.
Claude Opus 4.6 is good, but still a long way from where @AnthropicAI and the Vibe Coding community need it to be.
Here is where things stand with all the throttling and Opus nerfing debacle:
1. The usage limits are brutal.
I hit the $200 Max plan weekly limit within 2 days of heavy usage.
I have never hit the same even with 7-days of heavy usage.
On the 12th April, Claude was essentially unusable for a few hours. As a gesture, Anthropic offered $200 in bonus credits, presumably because they knew the throttling was aggressive and Claude was playing dumb.
2. The bonus credits are not the relief they sound like.
Today, Claude spent nearly $100 attempting to fix a single bug.
I am now on the second bug, and it will likely exhaust the remaining credits just during the audit and planning phase.
So until my limit resets, my options are to sign up for a second plan or sit idle on my project entirely.
3. The "fraction of the cost" argument does not hold up.
Yes, Anthropic gives access to more compute at a fraction of the API cost. But it conveniently ignores the vendor lock-in and infrastructure adoption happening on Anthropic's terms.
If the goal is to throttle models and reduce limits, the honest move would be to ship a model that actually justifies the cost per task, not one that burns $100 fixing a lightweight bug.
Instead, this looks like softening users up for a Mythos launch priced at 2x Opus or more. That is what all this hype is quietly building toward.
Stop performing transparency while withholding it.
Stop acting above criticism while paying customers hit walls mid-project.
If you want loyalty, earn it by actually caring about the people funding your roadmap.
All of these daily releases mean nothing if the underlying core behind it is not even usable.
@trq212@DarioAmodei
Introducing Claude Design by Anthropic Labs: make prototypes, slides, and one-pagers by talking to Claude.
Powered by Claude Opus 4.7, our most capable vision model. Available in research preview on the Pro, Max, Team, and Enterprise plans, rolling out throughout the day.
I agree with @Charles_SEO, there's two ways to look at it:
1. Discovery
2. Citation
From GEO perspective, it is good to properly cover the theory fan-out properly.
However, a major part of AI citations is also being discovered first in traditional search.
If you spam your own domain with content cannibalism, you actually risk killing your own citations as an effect of losing discovery in traditional search.
So, you want to cover for fan-out, but the proper way to go about it would be, cover the topics well with depth on your website, instead of cannibalising.
Cover the fan-out with a set of trusted sources, something like a target guest post campaign on good sites.
It actually helps you in two ways:
1) Helping you cover your fan-out properly.
2) Provide additional citations/sources build trust.
Here's some math for those having a hard time understanding it.
In the past one-week, I have paid nearly $239 in my monthly subscription, and about $200 in extra usage, and yet, I can't use it now until my weekly limit resets on Sat morning.
There has to be a cost to value proposition in there somewhere 😀
Anthropic should stop beating around the bush and commit to releasing Mythos, instead of building unnecessary hype around it.
Claude Opus 4.6 is good, but still a long way from where @AnthropicAI and the Vibe Coding community need it to be.
Here is where things stand with all the throttling and Opus nerfing debacle:
1. The usage limits are brutal.
I hit the $200 Max plan weekly limit within 2 days of heavy usage.
I have never hit the same even with 7-days of heavy usage.
On the 12th April, Claude was essentially unusable for a few hours. As a gesture, Anthropic offered $200 in bonus credits, presumably because they knew the throttling was aggressive and Claude was playing dumb.
2. The bonus credits are not the relief they sound like.
Today, Claude spent nearly $100 attempting to fix a single bug.
I am now on the second bug, and it will likely exhaust the remaining credits just during the audit and planning phase.
So until my limit resets, my options are to sign up for a second plan or sit idle on my project entirely.
3. The "fraction of the cost" argument does not hold up.
Yes, Anthropic gives access to more compute at a fraction of the API cost. But it conveniently ignores the vendor lock-in and infrastructure adoption happening on Anthropic's terms.
If the goal is to throttle models and reduce limits, the honest move would be to ship a model that actually justifies the cost per task, not one that burns $100 fixing a lightweight bug.
Instead, this looks like softening users up for a Mythos launch priced at 2x Opus or more. That is what all this hype is quietly building toward.
Stop performing transparency while withholding it.
Stop acting above criticism while paying customers hit walls mid-project.
If you want loyalty, earn it by actually caring about the people funding your roadmap.
All of these daily releases mean nothing if the underlying core behind it is not even usable.
@trq212@DarioAmodei
I am having a hard time using it on the $200 plan lol
This week, in the past one-week, I have paid nearly $239 in my monthly subscription, and about $200 in extra usage, and yet, I can't use it now until my weekly limit resets.
There has to be a cost to value proposition in there somewhere 😀
https://t.co/EE4L4pIl9n
@Prathkum I couldn't get past the welcome screen of Cursor.
I have shipped a fully working app with tons of features with Claude.
Make of that what you can.
- Non-Coder Perspective 🤷♂️
Yesterday was the worse Claude Opus 4.6 day for me in recent days.
I probably used most amount of tokens I ever have, yet shipped the least!
In fact, I am still sanitizing a lot of mess Claude created yesterday.
CLAUDE OPUS 4.6 IS NERFED.
BridgeBench just proved it.
Last week Claude Opus 4.6 ranked #2 on the Hallucination benchmark with an accuracy of 83.3%.
Today Claude Opus 4.6 was retested and it fell to #10 on the leaderboard with an accuracy of only 68.3%.
A 98% increase in hallucination.
https://t.co/ttnnwBYerW just confirmed that Claude Opus 4.6 has reduced reasoning levels and is nerfed.
You can now enable Claude to use your computer to complete tasks.
It opens your apps, navigates your browser, fills in spreadsheets—anything you'd do sitting at your desk.
Research preview in Claude Cowork and Claude Code, macOS only.