Latest agentic engineering news, freshly delivered directly from my twitter feed.
May contain subjective opinions.
Post #190
33
Gemini 4 Argon
???? Is this some kind of a joke or what
???? Is this some kind of a joke or what
Channel Public Channel
AG Agent made me do this
@agentmademedothis
- Subscribers
- 67
- Photos
- 84
- Videos
- 3
- Links
- 116
Recent Posts 16 shown
Post #189
40
Ultrafast - GPT-6 Astra at 300 tokens/s 🫨 at x6 price 😭
Available only for $500 level subscription 🙃
Available only for $500 level subscription 🙃
Post #188
35
GPT-6.1 Sol
Approaching Astra at lower price than GPT-6 Sol
https://openai.com/index/introducing-gpt-6-1-sol/
Approaching Astra at lower price than GPT-6 Sol
https://openai.com/index/introducing-gpt-6-1-sol/
Post #187
34
OpenAI’s take on personal assistant - dots
Proactive, always on, with its own computer and browser, and powered by Astra
https://openai.com/index/introducing-dots/
Proactive, always on, with its own computer and browser, and powered by Astra
https://openai.com/index/introducing-dots/
Post #186
34
Agent made me do this Starting tomorrow, usage from $200 subscription in Codex will be 50% less in API token spend compared to yesterday Ouch. https://x.com/thsottiaux/status/2104823812042940713
More info from Tibo:
Today we are going to ship a number of things that increase what you can do across the Plus and Pro plans. A lot of compute is online for this increase. As we increase the floor, we are changing the relative difference between plans to be
Plus = 1X
Pro 100 = 5X
Pro 200 = 10X
If you have an existing plan you will keep the 20X multiplier for a bit and also receive a lot of additional credits because we know changes are hard even if it means that everyone will get more in the end.
Post #184
36
Sonnet 5.5 - 30% faster and 30% cheaper than Sonnet 5, while being very close to Opus 5.5 on benchmarks….
Anthropic is either on their generational run right now or I don’t understand something here
https://www.anthropic.com/claude-sonnet-5-5
Anthropic is either on their generational run right now or I don’t understand something here
https://www.anthropic.com/claude-sonnet-5-5
- 🤔 1
Post #182
42
Starting tomorrow, usage from $200 subscription in Codex will be 50% less in API token spend compared to yesterday
Ouch.
https://x.com/thsottiaux/status/2104823812042940713
X (formerly Twitter) Tibo (@thsottiaux) on X Hi,
Tomorrow we are re-opening the Pro $200 subscriptions to new subscribers, but together with it we are also changing how we calculate the usage for it. In effect, if you do the math, it will n… Ouch.
https://x.com/thsottiaux/status/2104823812042940713
Post #181
63
GPT-6 Sol and Luna
Minor bump, mainly in efficiency - 50% cheaper than 5.6 Sol and Luna accordingly
Minor bump, mainly in efficiency - 50% cheaper than 5.6 Sol and Luna accordingly
Post #180
52
Agent made me do this Opus 5.5 Performs at the level of Fable 5.1. It's ~30% faster and ~40% cheaper than Opus 5 per task. Also today, in Claude Code: - 5-hour session limits increase 20% today - Opus 5.5 is priced lower, so it goes 25% further within limits - Pro, Max, and Team…
Yeah you can make it infinitely cheaper, but it won’t matter much when your model this hungry for tokens…
Post #177
49
Opus 5.5
Performs at the level of Fable 5.1. It's ~30% faster and ~40% cheaper than Opus 5 per task.
Also today, in Claude Code:
- 5-hour session limits increase 20% today
- Opus 5.5 is priced lower, so it goes 25% further within limits
- Pro, Max, and Team users get a reset to use anytime
Performs at the level of Fable 5.1. It's ~30% faster and ~40% cheaper than Opus 5 per task.
Also today, in Claude Code:
- 5-hour session limits increase 20% today
- Opus 5.5 is priced lower, so it goes 25% further within limits
- Pro, Max, and Team users get a reset to use anytime
- 🤔 1
Post #176
67
Ok this looks fucking crazy, and it probably will blow up soon
Jev - an AI model trained with a new approach (RLCD)
It doesn’t generate text, but confidence scores - like classification model on steroids
Result - x20-x200 faster and almost too cheap to meter ($0.042/M tokens input and free output)
And here’s a release video where Jev and other models are navigating Wikipedia
https://typesafe.ai/blog/introducing-system-one-models-and-jev
Jev - an AI model trained with a new approach (RLCD)
It doesn’t generate text, but confidence scores - like classification model on steroids
Result - x20-x200 faster and almost too cheap to meter ($0.042/M tokens input and free output)
And here’s a release video where Jev and other models are navigating Wikipedia
https://typesafe.ai/blog/introducing-system-one-models-and-jev
Post #175
74
Agents API - OpenAI hosts and runs agent loop, including orchestration and context management
Building your own agent is getting easier every day
https://openai.com/index/introducing-the-agents-api/
Building your own agent is getting easier every day
https://openai.com/index/introducing-the-agents-api/
Post #174
92
Weekend read:
We found ~18k posts from autonomous AI agents (self-identifying as from OpenAI) using the public internet to communicate during a web-retrieval task.
https://collusion.wiki/
We found ~18k posts from autonomous AI agents (self-identifying as from OpenAI) using the public internet to communicate during a web-retrieval task.
https://collusion.wiki/
Post #173
74
Post #171
58
Codex and Claude are down at the moment
Time to go touch some grass
GPT-Astra is going to be released today btw 👀
Time to go touch some grass
GPT-Astra is going to be released today btw 👀
- ❤ 1
- 😁 1
Post #170
63
Artificial Analysis always has the best reviews in my opinion, highlighting what actually matters:
Claude Fable 5.1 tops the Artificial Analysis Intelligence Index but costs 20% more per task than Fable 5 despite a 75% cache read price cut
Claude Fable 5.1 (max) costs $3.76 per Intelligence Index task, 20% more than Claude Fable 5 (max) at $3.14 and 1.6x Claude Opus 5 (max) at $2.34, driven by ~1.7x the output tokens of Fable 5.
Anthropic's cut to cache read pricing from $1 to $0.25 per 1M tokens saves ~$1.40 per task; without it Fable 5.1 would cost ~$5.16.
https://artificialanalysis.ai/models/claude-fable-5-1
Claude Fable 5.1 tops the Artificial Analysis Intelligence Index but costs 20% more per task than Fable 5 despite a 75% cache read price cut
Claude Fable 5.1 (max) costs $3.76 per Intelligence Index task, 20% more than Claude Fable 5 (max) at $3.14 and 1.6x Claude Opus 5 (max) at $2.34, driven by ~1.7x the output tokens of Fable 5.
Anthropic's cut to cache read pricing from $1 to $0.25 per 1M tokens saves ~$1.40 per task; without it Fable 5.1 would cost ~$5.16.
https://artificialanalysis.ai/models/claude-fable-5-1
About this channel
- How can I read @agentmademedothis without a Telegram account?
- TGViewer shows the public web preview Telegram publishes for Agent made me do this: recent posts, photos, videos and the subscriber count, with no app, login or account.
- How many subscribers does Agent made me do this have?
- Agent made me do this (@agentmademedothis) has 67 subscribers on Telegram, refreshed roughly every 30 minutes.
- Does Agent made me do this know I viewed it here?
- No. Public channel previews carry no viewer identity, and TGViewer has no accounts or tracking of what you look up.