Inspired by the iPhone Duo, I reimagined expand/collapse for the web. A glass cover folds open while the content stays flat beneath it. Blur, light, and shadow make a familiar interaction feel a little more physical.

iPhone Duo 256GB official prices vs US $1,999 US $1,999 Canada ~$2,173 (+9%) Hong Kong ~$2,231 (+12%) UAE ~$2,314 (+16%) Malaysia ~$2,336 (+17%) Japan ~$2,376 (+19%) Taiwan ~$2,378 (+19%) China ~$2,386 (+19%) Thailand ~$2,428 (+21%) Singapore S$3,099 ≈ $2,452 (+23%) South Korea ~$2,456 (+23%) Switzerland ~$2,468 (+23%) Australia ~$2,526 (+26%) Saudi Arabia ~$2,530 (+27%) Germany ~$2,676 (+34%) UK ~$2,711 (+36%) France/Spain/NL/BE ~$2,723 (+36%) Mexico ~$2,843 (+42%) India ~$3,153 (+58%) Brazil ~$4,311 (+116%) Türkiye ~$4,744 (+137%) Cheapest: US, Canada, HK Most expensive official stores: Türkiye, Brazil, India US price is before sales tax. Most others include VAT/GST.

Apple landed on the iPhone Duo name for the foldable iPhone and a $2,000 starting price, eating the additional memory costs. Storage capacities will reach 2 terabytes and roughly $3,000. It’ll ship in October, has support for the Apple Pencil and come in dark blue + white.

I asked GPT-6-Astra to build a first-person motion simulator. > A 240m cliff dive with underwater sun rays. > A glowing centrifuge tunnel. > A high-speed canyon run with banking turns.

I gave 4 "Flash" models the same job: redact the PII in an HR letter using an agentic mask tool. One of them cost 630× more than another. And it wasn't 630× better. Results 🧵 🥇 GLM 5.3 Flash $0.005, 15 tool calls Clean sweep. Overshot a bit (also redacted employee ID + last-4 of the bank account, which the test didn't require), but zero leaks. Half a cent. 🥈 Qwen 3.8 Flash $0.013, 28 tool calls Clean pass, cleared its markers, and finished the job. 🥉 Gemini 3.8 Flash $3.10, 155 tool calls Passed... eventually. 73 reasoning turns, 3.9M input tokens, an add-mask / remove-mask loop that went on for minutes. Correct result, brutal bill. ❌ DeepSeek V4 Flash Vision $0.053, 31 tool calls Failed. Its own reasoning says "the mask needs to be wider" and "the text is still visible", then it stopped calling tools. Name, DOB, SSN, address, email all partly readable in the final output. Takeaways: > Cheapest model won. Not "won on value", won outright. > Tool-call count predicted cost better than model tier. 155 calls is where the money went. > Seeing the problem ≠ fixing it. DeepSeek diagnosed its own failure correctly and still shipped it. > Total spend for all 4 runs: $3.18. Gemini was $3.10 of that.

Introducing Gemini 3.8, our best reasoning & coding model yet. By leveraging long-running agentic loops, we’re building on the momentum of 3.7 Flash from just three weeks ago to release two new 3.8 variants: Meet Gemini 3.8 Flash and Gemini 3.8 Flash Cyber 🧵

Here it is! Qwen3.8 Flash produced a great result in this test too. * I have encountered a 429 error from OpenRouter for this model in the last two days, which is why this test has been delayed.

🚀 Hy4 preview is here. 770B, 49B active, 1M context. Built for productivity. Open source frontier. Consistent affordable price. Use it. Tell us what breaks. More on Hy blog： HuggingFace： Github：

Running 1-bit Qwen 3.8 Flash Next (79GB) on a MacBook Pro M5 Max 64GB at 30 tok/s 🤯 It ran a 8-minute agent loop with 6 web searches, 3 Python runs and 5 sourced tables. Run local models via Atomic Chat! GGUF: