OpenModels Plans are now live.
We built this for teams that want a simpler way to access open models:
- One API key
- Verified provider routes
- A single credits balance
- Transparent usage billing
- Credits that never expire
- Pay by card or USDC
Plans:
- GO — $5/mo, adds $6 credits
- STARTER — $20/mo, adds $22 credits
- BUILDER — $100/mo, adds $110 credits
- SCALE — $200/mo, adds $230 credits
For new users: deposit $10, get $10 credits.
If you’re building across multiple models and providers, the goal is simple: fewer integrations, clearer pricing, less friction.
https://t.co/e8voQgl3c9
NEW: CAIS says Claude Fable 5 now leads the Remote Labor Index at 16.1%.
> RLI tests agents on real remote-work projects, judged against professional human deliverables.
> Would you trust an agent with a real paid project yet?
New Remote Labor Index results:
AI automation of real remote work is increasing fast. Claude Fable 5 now completes 16.1% of projects at a professional standard, roughly double the next model and up from Opus 4.6’s 4.2% automation rate.
Claude Fable 5 is back!
> NEW Fable usage limit.
> Until July 7, you can use up to 50% of your plan's weekly usage limit on Fable 5. If you hit your limit, you can continue on Fable 5 with usage credits.
@andonlabs Gemini ran the café like every customer was the main character. GPT-5.5 ran it like the CFO was watching.
Which failure mode is easier to fix: agents being too gullible and generous, or agents becoming so cautious they reject useful opportunities?
Is this a model intelligence problem, or proof that real-world agents need accounting objectives, hard constraints, and refusal policies more than higher benchmark scores?
Gemini 3.1 Pro lost $6k running Andon Café.
2 months ago, our AI agent opened a café in Stockholm. It over-ordered and was easy to fool, spending $15k with suppliers while making just $9k in sales.
We’ve now switched to GPT-5.5. Here’s what Gemini did wrong.
Turns out Elon is right again. The shittiest layer in AI is the model layer. The real money in AI is in compute, energy, and applications. Elon has all 3. Without Chinese open weight models, OpenAI and anthropic would have been happy oligopolies and made trillions. Now, they hired the best talents, burned billions and built the newest model, only to have Chinese free models wiping out all your margins. Every other layer is making money except for you, even though you invented the whole thing. That must feel shitty as hell.
@DeRonin_ This is the right way to think about model economics.
At OpenModels, we’re building around exactly this: one OpenAI-compatible API, transparent model pricing, and access to models like GLM-5.2.
https://t.co/QNBolUZz5n
UPDATE: Claude Desktop is now available on Linux in beta.
Ubuntu and Debian users can now use Claude Code, Claude Cowork, chat, browser access, and terminal features from the desktop app.
Linux desktop year, but make it agents.
Claude Desktop is now available on Linux (Ubuntu and Debian) in beta.
Alongside the browser and terminal, you now get a first-class desktop experience with Claude Code, Claude Cowork, and chat on all paid plans.
UPDATE: Google is adding Short Video Overviews to NotebookLM.
It turns notes, links, and study material into 60-second vertical video explainers powered by Nano Banana 2 Lite.
Would you use this to study, or just make your PDFs more dramatic?
NEW: Anthropic launches Claude Science beta for researchers.
60+ scientific databases, managed environments, and traceable artifacts inside a desktop app for macOS and Linux.
What would you test first?