惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

Vercel News
Vercel News
奇客Solidot–传递最新科技情报
奇客Solidot–传递最新科技情报
Apple Machine Learning Research
Apple Machine Learning Research
T
Tailwind CSS Blog
让小产品的独立变现更简单 - ezindie.com
让小产品的独立变现更简单 - ezindie.com
人人都是产品经理
人人都是产品经理
V
V2EX
量子位
Last Week in AI
Last Week in AI
Jina AI
Jina AI
博客园 - 【当耐特】
爱范儿
爱范儿
宝玉的分享
宝玉的分享
OSCHINA 社区最新新闻
OSCHINA 社区最新新闻
Hugging Face - Blog
Hugging Face - Blog
博客园 - 三生石上(FineUI控件)
有赞技术团队
有赞技术团队
小众软件
小众软件
IT之家
IT之家
博客园_首页
博客园 - 聂微东
S
SegmentFault 最新的问题
阮一峰的网络日志
阮一峰的网络日志
博客园 - 叶小钗

The New Stack | DevOps, Open Source, and Cloud Native News

Agentic development hinges on verification. For cloud-native software, that is a runtime problem. AI agents need infrastructure: Why Europe’s regional cloud strategy matters Transform your AI coding agent into a deterministic Java Spring expert WeAreDevelopers is coming to the US to give unsung developers a bigger voice Cleaner AI training data, fewer bugs: Sonar’s SonarSweep explained Observability overload is drowning engineers Google’s DiffusionGemma is 4x faster than its other Gemma models Fable 5: Guardrails and burn rate are annoying users, who say it’s still better than Opus 4.8 The Anthropic leader who built Claude Code says he ditched prompting — now he just writes loops. AWS can now mathematically prove your VMs are isolated Microsoft pulled 73 GitHub repos after malware attack — but still won’t say who’s compromised Databricks wants to kill the “email me a file” problem for AI agent skills Ramp bets forward deployed engineers can do what off-the-shelf finance AI can’t Git real: AI agents aren’t just for solo developers anymore Anthropic launches Claude Mythos/Fable 5, but you better try it soon Spring is 23 years old. AI just made it a security emergency. This AI agent startup ditched Anthropic for DeepSeek — and says it’s saving millions When your data model is the bottleneck: lessons from Medium’s feature store How long before we stop reading the code? The tokenmaxxing party is over, and Revenium is mopping up How AI is solving the memory crunch it created Microsoft’s pitch to enterprises: Ditch Azure Repos for GitHub, despite its rocky reliability record Claude Code’s biggest upgrade yet ran 5 agents at once — here’s what happened Why Anthropic just doubled Claude Cowork limits at no charge For years, Apache Cassandra handed this work to your team — 6.0 takes it back “A dangerous combination”: The 2 factors that can “corrupt” AI agent workflows With Foundry, Microsoft bets the enterprise AI battle is about reliability, not capability Microsoft unlocks Visual Studio for developers left behind by its own AI AI teams now deploy 1,000 times a month. Your pipeline wasn’t built for that. Microsoft just made the agent runtime free — and kept everything around it
Google pushes Pro, Ultra, and free users from open-source...
Meredith Shubel · 2026-05-27 · via The New Stack | DevOps, Open Source, and Cloud Native News

Last week at Google I/O, the company announced the beginning of the end for Gemini CLI — that is, for everyone save enterprise users and those with API keys. Starting June 18, many users will lose access to Gemini CLI, Gemini Code Assist IDE extensions, and Gemini Code Assist for GitHub. 

Instead, developers can shift to Antigravity CLI, a closed-source platform that lacks features and is already frustrating some developers with usage limits. 

But why? 

In its announcement blog post, Google says Antigravity CLI comes in response to the community’s shifting needs: “Listening to your feedback made one thing clear: we can serve you best by pouring our energy into a single product built for today’s multi-agent reality.” 

That product, Antigravity CLI, is a “premier agent-first development platform” that includes a server-side harness and a new terminal experience. Google says Antigravity will still enable users to get quick answers, scaffold and build new projects, and provision cloud infrastructure, but it’ll be snappier. Plus, Antigravity can orchestrate multiple agents in the background for complex tasks and make sure improvements to the core agent are universally applied, thanks to a new unified architecture. 

But going by the first waves of feedback from Gemini CLI users, it appears many think they’re getting the short end of the stick. 

It has fewer features, at least for now

“There won’t be 1:1 feature parity right out of the gate,” says Google. What it doesn’t say is when, or if, that parity will ever come. 

“There won’t be 1:1 feature parity right out of the gate.”

In the meantime, developers can use what Google dubs Gemini’s “most critical features” via Antigravity plugins: Agent Skills, Hooks, Subagents, and Extensions. 

It’s not open source

Unlike Gemini, Antigravity isn’t open source. While the Gemini GitHub page is dense with code and a list of hundreds of contributors, Antigravity’s is bare in comparison. 

Developers do not seem enthused. As one Redditor notes: “I don’t see any indication that Antigravity cli will be open source, I had all kinds of custom layers on top of gemini cli that I will be sad to lose. And I am quite anxious about usage limits.” 

It may be effectively more expensive

Other developers aren’t merely anxious about usage limits; they’re already running into issues. The GeminiCLI Reddit thread is full of frustrated comments from Antigravity users saying the new platform makes them hit usage limits faster, which may effectively render the new platform more expensive than was Gemini. 

“I can tell you quota is VERY LOW with Antigravity CLI, just tried do design a couple of screens in kotlin, and run out of tokens,” says one Redditor. Another adds, “even with pro I got the usage limit in just 6 to 7 prompts, this is insane, earlier I used to make whole projects with gemini cli with only 13% quota reached.”

Still more chime in with reports of their own debacles, one questioning whether Antigravity was even ready for a rollout, “Something’s wrong with the limits; they’re catastrophically low even with a subscription. The documentation is very poor. They shouldn’t have released it in this state.” 

The trouble with usage limits seems to be the common story, and it may end up pushing some users to jump ship. “if they don’t refresh every day I will cancel my google sub and use codex or claude code. The pro tier is a joke if Cli quota is gone,” writes another Redditor. 

“Just switched to antigravity cli but going back to gemini cli while I can, my tokens went crazy fast,” says one more. 

But the window to switch back is closing fast. 

Everyone’s out of luck, except enterprise users and those with API keys

On June 18, Google AI Pro, Ultra, and free users will lose access to Gemini CLI, Gemini Code Assist IDE extensions, and Gemini Code Assist for GitHub — but enterprise users and those with API keys are safe. 

For users with paid Gemini and Gemini Enterprise Agent Platform API keys, Gemini CLI will still be accessible. Similarly, for those using Gemini CLI or IDE extensions via a Gemini Code Assist Standard or Enterprise license and for those using Gemini Code Assist for GitHub via Google Cloud, access will be unchanged. 

Is this the way things are now?

“[N]ot a long time ago I was eager for some AI-related news, cause I expected some new breakthroughs. Now every announcement is [basically] ‘we’re making AI more expensive,’” laments a Redditor. 

“Now every announcement is [basically] ‘we’re making AI more expensive.'”

If they’re sour about so many fellow Redditors saying they’re hitting usage limits faster with Antigravity than they did with Gemini, this user may be similarly sullen about Anthropic splitting billing and pushing Agent SDK into separate credit pools and GitHub moving Copilot to usage-based billing, other moves in the growing trend of climbing AI costs. 

YOUTUBE.COM/THENEWSTACK

Tech moves fast, don't miss an episode. Subscribe to our YouTube channel to stream all our podcasts, interviews, demos, and more.

Created with Sketch.