In partnership with

Hey {{first_name | there}},

Claude Code is powerful, no doubt. But lately, its quality has degraded. Not to mention the weekly limits that eat half your workday.

So here are some Claude Code alternatives you can try instead. 

They're not as good as Claude Code at its best, but they come close. And most importantly, they actually let you keep working when Claude cuts you off.

Here's the lineup:

1. OpenCode
A solid open-source harness that works with any model. You can pair it with open-weight models like GLM and Kimi. This sits somewhere between Sonnet and Opus in quality.

Just one heads up: It's a harness, not a model. You still need to bring your own model to plug into it.

How to set it up:
curl -fsSL https://opencode.ai/install | bash
Or npm i -g opencode-ai / brew install anomalyco/tap/opencode.

Once installed, run opencode in your terminal. Use /connect to authenticate a provider (Z.AI, OpenAI, Anthropic, etc.), /models to pick your model, and Tab to switch between Build and Plan modes.

5.3-codex is about like Sonnet, maybe slightly better. 5.4 is close to Opus. You won't hit limits as easily as with Claude. It's pretty good if you're working on simpler coding projects like CRUD apps, small scripts, or just bug fixing.

The tradeoff: Claude Code usually nails it in the first try. Codex would need 2-4 iterations to get there.

How to set it up:
npm i -g @openai/codex Or brew install --cask codex. 

Run codex in your project folder and sign in with your ChatGPT account (Plus/Pro/Business plans include Codex access) or an API key. Use /model in-session to switch between GPT-5.4 and GPT-5.3-Codex.

The VS Code fork that basically everyone used before Claude Code existed. It still holds up. Cursor Pro gives you unlimited Tab completions and a $20 monthly credit pool for premium models (Claude Sonnet, Opus, GPT-5.4). Auto mode is unlimited and picks the best model for each task, which stretches your budget a lot further than manually choosing Opus for everything.

Limitation: Heavy users burn through the credit pool fast, especially in Agent mode. Watch the usage meter or you'll get a surprise bill.

How to set it up:
Download from cursor.com, sign in, and import your VS Code settings when prompted. That's it.

Open any project folder with Cmd/Ctrl+O, hit Cmd/Ctrl+L for chat or Cmd/Ctrl+I for the Composer (multi-file agent). If you want to bring your own API key to avoid the credit pool, go to Settings, Models and paste it in.

Pi is a minimal terminal coding harness. You build or install what you need.

It works with 15+ providers out of the box. Anthropic, OpenAI, Kimi, OpenRouter, Ollama, and more. Switch models mid-session with Ctrl+L. So when Claude hits its limit, you press one shortcut and keep working on Kimi or GPT. No config file edits, no restarts.

Limitation: It's not plug-and-play. You'll spend time setting it up or installing community packages to get features Claude Code has by default. If you want something that just works, this isn't it.

How to set it up:
npm install -g @mariozechner/pi-coding-agent
Run pi in your project folder. Configure providers with pi config. Install community packages with pi install npm:@foo/pi-tools.

🛠️ AI Tools Worth Checking Out 

  • Recall — Save articles, videos, and notes, then chat with your own knowledge base

  • Remio — Turn notes, podcasts, and thoughts into a searchable personal AI assistant

  • Whacka — Build simple apps directly from your phone using AI

  • TranslaBuddy — Translate conversations in real time during calls or meetings

AI News of the Day

What’s going on: Sneaker brand Allbirds basically gave up on shoes after selling its business for ~$39M and is now pivoting fully to AI. They’re rebranding to “NewBird AI” and using fresh funding to buy GPUs and build AI infrastructure. This comes after years of losses and a massive drop from their $4B peak valuation.

Why it matters: This is peak AI hype. When struggling companies can just rebrand into “AI companies,” you know the market is hot. Some of these pivots will work, but a lot are just narrative resets to attract investors. The bigger takeaway: AI isn’t just a tech trend anymore, it’s becoming the default direction for any company trying to survive.

What’s going on: Anthropic is starting to roll out ID verification for Claude. Some users are being asked to upload a government ID and even do a selfie scan. They say it’s to stop abuse, bots, and people bypassing limits, but it’s not super clear who gets flagged and when.

Why it matters: People are pissed. Not just because of verification, but because this is a big shift. AI tools were kinda anonymous till now, this changes that. If this sticks, using powerful AI might mean tying your real identity to it. Less privacy, less freedom to mess around, more control from companies. And once this becomes normal, every AI tool might follow.

What’s going on: OpenAI just backed out of a big data center deal in Norway (part of its “Stargate” infra push). 

Instead of building/owning that compute directly, they’re now planning to rent it through Microsoft. Basically, they couldn’t agree on terms with the provider, and Microsoft stepped in and took the capacity. This also comes right after OpenAI paused a similar UK project and started toning down its massive spending plans.

Why it matters: OpenAI was going all-in on building insane amounts of its own AI infrastructure, but now it’s pulling back a bit and leaning more on Microsoft again. Could mean costs are getting out of hand, or they’re being more careful before a potential IPO. Either way, it shows even the biggest AI players can’t just scale infinitely without hitting real-world limits (money, energy, deals).

What’s going on: Someone built a plugin for Claude Code that can extract a website’s design system. Things like colors, fonts, spacing, and components, and turn it into usable files (like Tailwind config, CSS variables, and a structured markdown Claude understands). 

You can then use that to recreate similar layouts faster, but it’s not a perfect clone and can struggle with things like heavy bot protection or complex apps.

Why it matters: This makes frontend work way faster, especially if you’re not great at design. Instead of starting from scratch, you can study and reuse patterns from existing sites. But it also lowers the barrier a lot; good design becomes easier to replicate, so standing out will matter more than ever.

Final Takeaway

AI is moving in two opposite directions at the same time. 

On one side, tools are getting insanely powerful and easier to use, like extracting designs or building apps in minutes. On the other, companies are tightening control with things like ID verification and infrastructure shifts.

Basically: more power, less freedom.

If you’re building or using AI, this is the phase where things start getting serious. The people who win won’t just use the tools, they’ll understand where all of this is heading and adapt early.

- Aashish

Become An AI Expert In Just 5 Minutes

If you’re a decision maker at your company, you need to be on the bleeding edge of, well, everything. But before you go signing up for seminars, conferences, lunch ‘n learns, and all that jazz, just know there’s a far better (and simpler) way: Subscribing to The Deep View.

This daily newsletter condenses everything you need to know about the latest and greatest AI developments into a 5-minute read. Squeeze it into your morning coffee break and before you know it, you’ll be an expert too.

Subscribe right here. It’s totally free, wildly informative, and trusted by 600,000+ readers at Google, Meta, Microsoft, and beyond.

Reply

Avatar

or to participate

Keep Reading