Name here

@burkeholland @github Rubber duck is god mode.

I ❤️ Opus, BUT 💰💰💰💰💰 I've been using Sonnet 4.6 (Medium) with Rubber Duck in Copilot because @Github reports ~75% of Opus perf w this strat. I'm loving it. Fast, has not let me down yet and was able to fix a bug that even Opus could not. github.blog/ai-and-ml/gith…

@spikednn @RnaudBertrand Folks here are talking about the api. What real men use

@RnaudBertrand Not even remotely true. Compare against the subscription which is what everyone uses and it’s pretty much the same price. You can do about 2B tokens a day on the $200 subs. Just enjoy the subsidies while they last.

Hard to calculate exactly without an input/output split but I did the math and for 831,962,136 tokens, Anthropic Opus 4.7 would cost: - 100% input (floor, unrealistic): 831.96M × $5/M = $4,159.81 - 90/10 (typical for coding agents like opencode — most tokens are codebase context re-fed each turn): $3,743.83 + $2,079.91 = $5,823.74 - 80/20 (more conservative): $3,327.85 + $4,159.81 = $7,487.66 - 50/50 (worst plausible case): $2,079.90 + $10,399.53 = $12,479.43 So that $10.57 DeepSeek bill would probably become roughly $5,000–8,000 on Claude Opus 4.7. In other words, DeepSeek is 500–700× cheaper, for similar-ish capabilities. Now you start to understand why Anthropic is worried...

@DanWahlin Rubber duck. Best of both worlds

I've been using GPT 5.5 heavily for the last few days in GitHub Copilot CLI and have been really impressed so far. Opus has always been my go to, but I feel like 5.5 is generating better results overall. Anyone else using it a lot? Thoughts...good or bad?

@merill @DanWahlin Codex app > Copilot via CLI and GitHub app?

@DanWahlin Yup, I've been a Claude believer ever since I released the first Microsoft Graph MCP (lokka.dev) back in Feb last year. BUT for the last couple of weeks, I've 100% switched to Codex and it is so good. The Codex app is 😘

@wongmjane Ppl will do anything but take public transit

SF self-driving trio during rush hour: Waymo: overpriced Zoox: too far away Robotaxi: “High service demand” Anyways, I ended up taking a yellow cab instead 🚕

@KuittinenPetri @ai_for_success Dude read the docs you can BYOK cmon mannnnnn

@ai_for_success GitHub Copilot went from having the best value for money in AI assisted coding to having no value over just paying for API credits yourself (coming in 1st of June 2026). Absolutely mind-blowing! They should at least add cheap Chinese OSS models to their model list. 1/

I was going through GitHub Copilot pricing changes… this is wild. They quietly changed model multipliers and some of these jumps are insane. Opus 4.6 is 9x Opus 4.5 is 5x Opus 4.7 is 3.6x Sonnet 4.6 is 9x Sonnet 4.5 is 6x Gemini 3 Pro is 6x Gemini 3.1 Pro is 6x GPT 5.1 is 3x GPT 5.2 is 3x GPT 5.3 Codex is 6x GPT 5.4 is 6x

@zeeg GitHub Copilot does this. No need to mess with api and all the other crap.

Alright Anthropic, OpenAI: I need APIs that give me usage data. Granular. Per user. I need this in the same way its provided by every other company for all of time. What I dont want: "heres a $100,000 line vague line item of token spend".

Claude Code vs OpenCode github comments hit different this way (im so sorry)

@handotdev Hell yeah let us know in 2035 when you get it done 💪

an ai-native, high performance, never-down github alternative would go hard right now

@grhmc Doom scrolling

I mean

@mitchellh I’m saddened more that you’ve devoted so many days and years of your fleeting life to a stupid software on a lifeless computer. Good luck

Ghostty is leaving GitHub. I'm GitHub user 1299, joined Feb 2008. I've visited GitHub almost every single day for over 18 years. It's never been a question for me where I'd put my projects: always GitHub. I'm super sad to say this, but its time to go. mitchellh.com/writing/ghostt…

@theo @cassidoo @film_girl x.com/theo/status/20…

@cassidoo @film_girl I really hope yall can bounce back here. GitHub is the center of so much of the software world. Rooting for you guys 🫡

The team is blessed and stressed y'all, please be nice in your tweets 😅 github.blog/news-insights/…

@MetacriticCap If I were to guess they’ve already lost single digit billions on subsidization so far

GitHub Copilot will continue to be subsidized, to the extent your company has a discount to the listing price, like many companies do. That said, extremely disappointing. Microsoft doesn't need to pay a dime to OpenAI to serve their models, Microsoft is extremely rich, they should at least losing some mid-single digit billions of dollars here before giving up.

@JoshCaughtFire @fjzeit @GitHubCopilot Ah, I gotcha. Looks we aren’t aligned. Cheers!

@collin_mit93900 @fjzeit @GitHubCopilot No, you would pick the best tool for your org… which likely isn’t copilot, especially coupled with GitHub’s poor uptime. The only thing really valuable about copilot was it’s pricing structure, which is now changing to be like everyone else, hence no value

my one question for @GitHubCopilot is: if you are to now charge $39 a month for API credits does this mean that unused credit will accumulate?

@JoshCaughtFire @fjzeit @GitHubCopilot Okay so you’re gonna pay the same price and onboard like 4 tools to your enterprise for the same functionality?

@collin_mit93900 @fjzeit @GitHubCopilot Dozens of other agents let you do that using open router, cloudflare, etc. as central billing. Seemed the edge with Copilot was their request based billing, I don’t think their agent is that spectacular

@JoshCaughtFire @fjzeit @GitHubCopilot So you can use all models under one sub. Copilot cli, etc

@fjzeit @GitHubCopilot That’s what I don’t get, wtf am I paying monthly for? And if I’m paying API prices, why would I use copilot

@olvrgln Damn a GitHub disrupter wouldn’t let this happen to them

@RimaSukhadia Pre production? I thought they had it figured out? It’s 2026!

@collin_mit93900 No driver, both have safety monitors… this is pre-prod Check out other posts fo glossy,

Shimmering Golden wrap ✨ Tesla pre-prod CyberCab

@GergelyOrosz Look at the amount users on GH vs those other services? Are you dumb? Like actually an idiot?

I totally get that agentic workflows means that GitHub has a lot more load to deal with (10x or more) BUT so do a bunch of other infra startups seeing 10x or more load increase - may that be Vercel, Resend, Railway, Cloudflare, Linear etc. What makes it a lot more challenging for GH? Git? Because when I look to those other startups that likely see a similar % of load increase, reliability does not seem to suffer as much, as constantly as GitHub.