@firtoz They obviously want you to visit the site so they can show you ads or do upsells. Pretty sure such sites would want to ban all agents, not just Muse
really tried using Gemini Carbon, but its just no match for claude models yet. waaay too slow and token hungry.. code reviews are good but findings are often incorrect...
@QuinnyPig Also its S3 emulator is limited to 64MB of in-memory data only, with no option to increase its hardcoded limits. Pretty much useless for long running integration tests...
Hello you fine Internet folks,
Today we’re looking at AMD’s next-gen Zen 6 CPU core family. We’ll explore the changes from Zen 5 and what the differences between Zen 6 and Zen 6c tell us about AMD’s approach to microarchitecture.
Hope y’all enjoy!
https://t.co/55jF197k6R
Anthropic has released Claude Fable 5.2, taking the #1 spot on the Artificial Analysis Intelligence Index v5 at 65, led by gains in science, knowledge, and coding
At max effort it scores 65, 5 points ahead of Gemini 4 Carbon (high, 60) and 6 ahead of Claude Opus 5.5 (max, 59). Gains are driven by scientific reasoning, knowledge and FrontierCode 2.0, and roughly matching Opus 5.5 on AA-Briefcase and Terminal-Bench 4.0.
Priced at $8/$40 per 1M input/output tokens, a 20% reduction from Fable 5.1, Claude Fable 5.2 averages a Cost per Task of $26.5 with max effort, a ~2.6x increase compared to Opus 5.5 (max, $10.0) and ~2x compared to Fable 5.1 (max, $12.99)
Key takeaways:
➤ Claude Fable 5.2 leads 5 of the 9 evaluations in the Intelligence Index: At max effort, Fable 5.2 scores 65, ahead of Gemini 4 Carbon (high, 60), Claude Opus 5.5 (max, 59) and GPT-6 Astra (max, 57), and ~15 points above Claude Fable 5.1 (max, 49). It sets the highest scores on HLE-Diamond (80%, previous best 69% from Gemini 4 Carbon), AA-Omniscience v1.1 (67, previous best 54 from Gemini 4 Carbon), Terminal-Bench-Science (73%, previous best 65% from Gemini 4 Carbon), CritPt v2 hard subset (77%, previous best 72% from GPT-6 Astra) and FrontierCode 2.0 (49%, previous best 46% from Claude Opus 5.5)
➤ Less progress in agentic knowledge work and terminal coding: Claude Fable 5.2 improves on Fable 5.1 across capabilities, but sometimes only to Opus 5.5 levels. It reaches 1785 Elo on AA-Briefcase, within the confidence interval of Opus 5.5 (1800) and just behind Claude Sonnet 5.5 (1813). On GDP.pdf it reaches 24% all-pass against 26% for Opus 5.5. On Terminal-Bench 4.0 it scores 60%, level with Opus 5.5 and behind Kimi K3.1 Pro (67%), Gemini 4 Carbon (65%) and Sonnet 5.5 (64%). AutomationBench-AA is an exception, at 76% against 70% for Opus 5.5, narrowly behind Gemini 4 Carbon, 78%
➤ Lower efforts still take leading positions: With high and xhigh effort, Claude Fable 5.2 scores 62, ahead of every other model at any effort, using 98k output tokens per task against 173k for Opus 5.5 (max). At max effort Fable 5.2 costs ~2.6x more than Opus 5.5, but at lower effort settings it can outperform Opus 5.5 at lower cost per task
Other model details:
➤ Context window: 1 million tokens with image and text input
➤ Pricing: $8/$40 per 1M input/output tokens; cache writes $10 with 5-minute TTL, $16 with 1-hour TTL, and cache reads $0.20
➤ Effort settings: five (low, medium, high, xhigh, max). Intelligence Index evaluations were run at all five with Anthropic's default fallback enabled
Congratulations to @AnthropicAI on this impressive release!
Me: I think we can compress this data better. Think deeply from first principles. Dont tie yourself to a particular structure, format or algorithm.
Opus: I tried all the popular compression algos and their knobs, and some custom techniques. Here is a super obscure algorithm you never heard about. This beats everything on this particular distribution of data.
Me: 🤯🤯
Apparently OpenAI now blocks Claude from accessing all openai domain pages...
Asked Claude to compare OpenAI's approach with Cursor and Anthropic VCS and Claude cannot read it.
We built Pando, a new Git backend for our monorepo, to handle more pushes and fetches without changing how our engineers and agents use Git.
Pando cut push latency by ~80% and now serves nearly a petabyte of Git data a day. We can grow Git capacity alongside agent usage.
me: Opus make me a cache for these files
Opus: here is a cache. but i introduced a new intermediary format, that keeps only certain fields in compressed serialized form in memory and deserializes on demand.
I measured the perfect ratio of cpu/mem balance to keep only the right fields serialized.
me: 🤯
give Astra computer use access to Photoshop and tell it to make your favorite photo look like a frame from a chosen movie.. the results will be insane… it will even crop and fix the aspect ratio to match the movie. 🍿
@alexalbert__ I am still seeing the 5hr limit. it says for 'verified' accounts but there is no instructions on how to verify the accounts.... or to even check if my account is marked 'verified'
Codex (both CLI and desktop app) are so soo buggy and break so often. I think OpenAI need to use Claude Code Review product on their prs...
https://t.co/uMzNig6VCM
Sol in Ultrafast mode comes to screeching halt during the constant 'Compacting Context'.. like putting speed breakers in front of your F1 car.
OpenAI is the only major provider still stuck at sub 1M context...