Marius Buleandra

282 posts

Marius Buleandra

Marius Buleandra

@mariusbl

Applied AI @AnthropicAI | Ex. @ycombinator @AndurilTech

San Francisco, CA Katılım Haziran 2012
455 Takip Edilen656 Takipçiler
Marius Buleandra
Marius Buleandra@mariusbl·
What does autonomous work even mean? The biggest impediment for me has been the fact that I have to “crank” the model to continue in the direction that I want it to go. With Opus 4.7 there’s less “cranking”
Claude@claudeai

Introducing Claude Opus 4.7, our most capable Opus model yet. It handles long-running tasks with more rigor, follows instructions more precisely, and verifies its own outputs before reporting back. You can hand off your hardest work with less supervision.

English
0
0
2
74
Marius Buleandra retweetledi
Bubble
Bubble@bubble·
Bubble 🤝 Opus 4.7 @AnthropicAI brought us in before the release to test and send feedback, and we’re seeing a 5–10% improvement in cross-page visual consistency. Create an app with Bubble AI and you’ll notice the gains!
Bubble tweet media
English
3
8
20
1K
Marius Buleandra retweetledi
Anthropic
Anthropic@AnthropicAI·
New on the Anthropic Engineering Blog: How we use a multi-agent harness to push Claude further in frontend design and long-running autonomous software engineering. Read more: anthropic.com/engineering/ha…
English
318
927
6.7K
1.8M
Marius Buleandra retweetledi
Claude
Claude@claudeai·
Claude can now build interactive charts and diagrams, directly in the chat. Available today in beta on all plans, including free. Try it out: claude.ai
English
1.6K
3.5K
42.7K
11.8M
Marius Buleandra retweetledi
Mike Krieger
Mike Krieger@mikeyk·
More than a million people are now signing up for Claude every day. To everyone choosing to make @claudeai part of how they work and think: welcome.
English
163
221
3.8K
660.7K
Marius Buleandra retweetledi
cat
cat@_catwu·
We just launched /remote-control so you can continue local Claude Code sessions from your phone This is now rolled out to all Max users!
Noah Zweben@noahzweben

Announcing a new Claude Code feature: Remote Control. It's rolling out now to Max users in research preview. Try it with /remote-control Start local sessions from the terminal, then continue them from your phone. Take a walk, see the sun, walk your dog without losing your flow.

English
110
42
1K
68.3K
Marius Buleandra retweetledi
Claude
Claude@claudeai·
New in Cowork: scheduled tasks. Claude can now complete recurring tasks at specific times automatically: a morning brief, weekly spreadsheet updates, Friday team presentations.
English
982
1.7K
22.3K
8.2M
Marius Buleandra retweetledi
Lawrence Chen
Lawrence Chen@lawrencecchen·
Introducing cmux: the open-source terminal built for coding agents. - Vertical tabs - Blue rings around panes that need attention - Built-in browser - Based on Ghostty When Claude Code needs you, the pane glows blue and the sidebar tells you why. No Electron/Tauri. Just Swift/Appkit.
English
219
174
2.1K
356.6K
Marius Buleandra retweetledi
andrew gao
andrew gao@itsandrewgao·
"sf is dead" we're throwing a robot fight club party tomorrow night! + food, drinks, good vibes i'm no @michelleefang but... partiful below!
andrew gao tweet media
English
17
7
250
32.9K
Marius Buleandra retweetledi
Cyrus
Cyrus@cyrusnewday·
We just did the largest eval of AI for systematic reviews. Over 30,000 datapoints. Key findings: → 97% accuracy across screening and extraction → 98-99% reduction in time → AI caught relevant studies humans missed
Cyrus tweet media
English
2
18
70
21.5K
Marius Buleandra retweetledi
Kilo
Kilo@kilocode·
Opus 4.5 is here — we're breaking it down with the people who know it best. Join Kilo's DevRel team and Marius from Anthropic's Applied AI team for the scoop. Learn performance/prompting strategies and what it means for your workflow. Bring your questions. twitter.com/i/broadcasts/1…
English
0
6
24
3.4K
Marius Buleandra retweetledi
jeremy
jeremy@jerhadf·
one fact people won't realize immediately about opus 4.5: it's remarkably token-efficient. all-in it's often *cheaper* than sonnet 4.5 and other models for cost-per-task-success. glad sourcegraph is seeing this early in Amp! we find that opus 4.5 with medium effort is pareto dominant over sonnet 4.5 on swe-bench verified - outperforming it at 77.4% w/ 4x fewer tokens and 35% of the cost. it's smart about using only the optimal time & tokens it needs to solve the problem. sonnet is still great and this may not hold for other use cases, but many agentic coding use cases are seeing remarkable efficiency. curious what you all find! check out this interactive chart opus 4.5 made about it: preview.claude.ai/artifacts/050a… sticker prices aren't everything!
jeremy tweet media
Quinn Slack@sqs

Opus is worth it, and maybe cheaper all-in than Sonnet? Early rough non-representative numbers, for our own internal @AmpCode usage (avg cost $ per thread): - Sonnet 4.5: $1.83 - Opus 4.5: $1.30 (earlier checkpoint last week was $1.55) - Gemini 3 Pro: $1.21

English
4
8
173
37.3K

Keşfet

@AnthropicAI @claudeai @michelleefang @elonmusk @BarackObama @taylorswift13 @cristiano @BillGates