Claude Opus 5.5

@theoSep 25

Opus 5.5 got my ts-rust port working in 10 hours, and it has been grinding on performance for the last 24 hours. I've been working on this port on and off for about 4 months. I got to ~35% tests passing with GPT-5.6 Sol, and ~85% with GPT-6 Astra. Both models stalled hard once hitting those numbers, and ran in loops with no meaningful progress for days at a time. I've never had "enough Anthropic tokens" to try a Claude model on a port like this. Opus 5.5 feels practically unlimited, so I threw it a "/goal finish the port and make it faster". I can not believe how quickly Opus unblocked the work Astra was stuck on. It may have made this port an actually viable project. Absolutely mind blown right now.

@prathkumSep 25

Holy shit, Claude Opus 5.5 is now up to 40% cheaper than Opus 5. Same tasks. Same quality. Way less cost. Breakdown: Input tokens: $4/M (down 20%) Output tokens: $20/M (down 20%) Cache reads: $0.20/M (down 60%) Most of a coding session's cost comes from cache reads, so this alone can cut your bill by more than half. AI is slowly getting cheaper.

@exm7777Sep 25

Opus 5.5 is the best model launch of the year for me... it couldn't have been better the timing was impeccable, they basically made the GPT-6 releases pointless because this is what we're getting: - better, faster and cheaper than Fable 5.1 - with the same feeling Fable had, the one that made it so good to work with - cheap enough to actually run in prod - limits that aren't that bad anymore so a lot more people can afford it now, which genuinely opens up use cases that were out of reach before they put OpenAI on silent mode for a few days with this one

@weswinderSep 25

opus 5.5 is truly the best model we've ever had i can't wait for anthropic to nerf it

@thekitzeSep 25

opus 5.5 is a mindreader. im not even trying. no fancy prompts and skills and bs. i love how all of the ai bs infra we're building will be rendered useless with smarter models lol

@theoSep 24

Opus 5.5 just filed a pull request that changes the streaming behavior in T3 code. It's so cool that I can send off a prompt with a vague idea of what I want, and the result is a ready-to-merge pull request with a video demo of the changes directly in the PR. The PRs I make with AI are significantly better than the ones I used to make by hand. The future is awesome.

@rileybrownSep 24

The only way I can describe using Opus 5.5 is that with it, I glide.

@thekitzeSep 24

unnerfed opus 5.5 i will miss you next week and always remember you as the greatest model that i've talked to

@omarsar0Sep 24

Interesting results here. This is why I expect more agent workloads to run on blended models. Pareto 26.9 from @TheUnbiasedCo sends requests to several frontier and open models and keeps the best answer. In the new eval of 30 agent tasks, Pareto tied GPT-6 Astra for first place at about 1/3 the cost per successful task. It also finished tasks faster than DeepSeek V4 Pro and GLM 5.3 Flash.

@thezviSep 24

Opus 5.5 reaction thread. We all know it can do original music videos, but how is it at talking and coding and all the stuff we will actually be doing next week?

@itsolelehmannSep 24

Anthropic has no image or video model, yet Opus 5.5 just made me this 30 second animated film using nothing but 2,800 lines of code (yes, this is 100% Claude Code with no other AI tools, connectors, or reference images) i gave it one prompt: 4 seasons passing outside a train window, a cozy carriage, a cup of coffee on the table, Grand Budapest Hotel style and it literally came back with the finished .mp4 file ready in the chat so i asked it... how did you achieve this result when you have no image model?? here's what Claude actually does behind the scenes: 1. sets up free drawing software on your computer, the kind that turns written instructions into pixels 2. writes ~2,800 lines of code describing every object as shapes with exact coordinates: a tree is a brown trunk plus ~6 overlapping green circles, the coffee cup is a few ovals and curves 3. layers the scenery at different speeds, so telegraph poles whip past 140x faster than the mountains (which is what gives it depth) 4. renders a still of each season, looks at them, then fixes what looks off 5. animates it like a flipbook: it calculates where every object should be at each moment, then redraws the whole scene 900 times 6. builds the soundtrack the same way, as equations. a plucked string is a stack of sound waves that fade out. each rail click lands on the exact frame where the coffee ripples 7. stitches it all into the final video it even invented its own season transitions: a passing train sweeps spring into summer, tunnel turns autumn into winter, etc the creativity and attention to detail is getting pretty ridiculous

@jackfriksSep 24

TIME TO WAKE-UP AND VIBE WITH MY BOY OPUS 5.5 UNTIL LUNCH (listen to music while opus 5.5 does work i told it to do 100x better than its little useless brother opus 5)

@aakashguptaSep 24

Opus-5.5 is the model I have liked most since Opus-4.6. It’s the type of model that only comes out once a year or so: a clear leap above what was possible before, really great to talk to, and all around banger. A lot of people have been upset with how Anthropic is watermarking text, or pacing the frontier, or being all doom and gloom. But what they have built over there has been working. As someone who follows every new model release from every lab closely, I can confidently say they have had the best frontier model for most of the last 2 years. That’s an insane feat in such a competitive market. Dario is cooking. The team is cooking. And the revenue shows it. Another year of 10x, when they themselves thought they couldn’t do it. I firmly believe they could do $1T revenue next year and 10x again. The funny thing is I think the team at Anthropic may not know it. The only company who has briefly taken the #1 spot is OpenAI. I think they still have the power to take it longer-term. But they need to learn more from Anthropic. They need to release more and only announce when they can release to everyone. They need to reach parity with Claude Code’s harness. They need to actually focus on AI writing because Anthropic went the watermarking route so they have clear counter positioning. Whether they can do so remains to be seen. Google, Meta, and X are also horses in the race. But they have yet to come close to a model of Opus-5.5’s particular delightfulness. It mogs Fable-5.1 in 80% of tasks, and it’s like 20% the cost (all things considered). It’s fucking awesome.

@aakashguptaSep 24

Claude Opus 5.5 animated this in one Claude Code session. A Series C offer the recruiter calls $370K a year: $200K base, $30K bonus, $40K sign-on, $400K in RSUs. Say you stay 2 years. Spread the sign-on over those 2 years. Value the stock at 40% odds of an IPO, 5 years out, discounted 15% a year: $400K × 0.4 ÷ 1.15^5 ≈ $80K, half of it vesting while you're there. That's $270K a year. Do the math before you counter.

@theoSep 24

Anthropic was so mad about people distilling Fable that they decided to do it themselves with Opus 5.5

@vasumanSep 24

Opus 5.5 just dropped meaning you have one week to squeeze every ounce of intelligence out of it before it gets nerfed down to Haiku 3.5

@rauchgSep 24

ShellPerfBench. I'm really neurotic about the startup time of a new shell session. Opus 5.5 found a lot of really great optimizations other models missed. Anon, anytime you open a new terminal, you might be silently suffering. Tell your agent to optimize .𝚣𝚜𝚑𝚛𝚌 and friends

@petergyangSep 24

Opus 5.5 is: - Smart - Fast - Fun to talk to - Good at writing - Friendly on rate limits - Constantly surprising at just how much it is capable of

@alexfinnSep 24

Stop scrolling for a second. I just want you to pause for a moment. We live in the most incredible time to be alive in the history of this species Every other day a new revolutionary product drops that gives you more freedom, power, and ability to do ANYTHING you want LITERALLY every other day A couple weeks ago Astra gave you super intelligence. Today Opus gives you super intelligence at lightning speeds for dirt cheap. Soon Meta glasses will come out that let you talk to a super intelligent personal agent everywhere you go Every day technology makes your life better and better. Every day you're capable of accomplishing more. Every day you have the ability to help out your fellow man more than ever before For thousands of years nothing happened. But today, everything is happening Please just take a moment and realize how incredible, awe inspiring, violently beautiful this all is If you are one of the pessimists that haven't realized the incredible nature of what is happening right now, I beg you to view this world through a different lens Being alive right now is the greatest gift God has ever given us. I hope everyone realizes this soon enough

@omarsar0Sep 24

Don't sleep on using Jev-as-a-Judge for agent evaluation. This is one of the most impressive Jev use cases I have found so far. Jev is a natural fit as a Judge, but it doesn't mean you use it everywhere. Similarly, you shouldn't use frontier models for evals everywhere. I'm running lots of tests on this atm, but early results point to an optimized flow (balancing accuracy and cost) that combines Jev and frontier models. Concretely, use Jev in high-confidence situations, and escalate to a frontier model (GPT-6 or Opus 5.5) in low-confidence verdicts. Entire write-up coming soon. Let me know if you have questions as I build the full guide.