GPT

A researcher reports that attacker-controlled recipient names can insert phishing links into authenticated ChatGPT Trusted Contact invitations. The flaw could place a malicious link inside an email sent from OpenAI.

Users report GPT 5.6 Sol's selected 1M-token context window initially reset to roughly 258K after a message. A later user confirmation said the 1M setting was restored.

LLMJunky demoed Slopet League, a Rocket League-style prototype in Godot, saying Opus 5 made the map, physics, effects, and animations. He says GPT 5.6 Sol made car meshes and the build took about 1.5 days.

Omoore says ChatGPT Sites built a playable “chronically online” quiz game with images, interactive modules, levels, and an anonymized dashboard. Only 4% of tracked players reached the top tier.

Creators built hosted ChatGPT Sites from prompts, including a GeoGuessr-style game with ChatGPT auth, secure key storage, Google Maps keys, OpenRouter opponents, and budget limits. Another site used ChatGPT Images for a math game UI.

Goodside compared Claude Fable 5 Max puzzle generation with GPT-5.6 Sol Pro solving attempts. Sol solved a 150-Pokémon empty crossword but failed the 1,025-Pokémon version, with one reported success traced to the answer key.

Demos from Gokay and Meng To used video, object-view, and image references to generate HTML, Three.js, or Blender scenes. The builds paired Sol or Claude Code with tools including Reve, HyperFrames, and Blender MCP.

Min Choi collected early GPT-5.6 Sol builds spanning Blender scenes, After Effects automation, Three.js characters, UI cloning, SQL DOOM, and Claude Code tests. Meng To also published a video-to-HTML workflow.

Goodside tested GPT-5.6 Sol and Claude Fable 5 with fake handwriting, constrained-vocabulary prompts, and Ghost Font. Sol often answered unreadable inputs, while Fable more often refused or pushed back.

Goodside fed GPT-5.6 Sol and Claude Fable 5 binary noise and meaningless handwriting. Sol often invented hidden text, while Fable also failed on noise but more often pushed back on scribbles.

Meng To says Sol Ultra generated an HTML/React canvas with layers, inspector panels, media items, and real-time collaboration. The test used planning and subagents, but it burned through limits.

GPT-5.6 Sol Ultra became generally available, and early users tested it on cinematic sites, games, Blender scenes, and Seedance comparisons. Reports praised orchestration but noted mixed physics and one 2B-token overnight run.

Users reported the former Codex Superapp now appears as ChatGPT Work, with Codex inside the ChatGPT desktop flow. Early posts flagged broken sites support and confusing sidebar states.

GPT-5.6 Sol appeared in Figma Make and creator benchmarks against Fable 5 across design, games, and video. Testers praised browser persistence and design output, while noting higher token use and mode confusion.

OpenAI staff said the new ChatGPT Voice is powered by GPT-Live-1 for more natural conversations. The launch also refreshed the ChatGPT shader with Blender prototypes translated to Metal and WebGL using Codex.

Three builder threads shared reusable layers around model APIs: per-user usage gateways, audits for Gemini-enabled GCP keys, and config-driven routing that swaps providers without app rewrites. Wrapping rate limits, key scope, and model choice in one layer helps teams ship multi-user apps without scattering provider logic.

OpenAI’s shutdown post says the Sora app is going away and that work-preservation and API timing details are still coming. For creators, that removes a standalone video surface and is already colliding with complaints about the remaining video offer and Pro value.

Glass says its Mac editor can tap existing Claude, ChatGPT and Gemini subscriptions inside one coding workspace, avoiding separate API keys and usage meters. Compare the flat-subscription workflow against Cursor-style billing before you move a product build.

Revise debuted as a document-first editor that keeps rewriting, proofreading, and generation inside the page instead of a separate chat window. Writers should compare the inline workflow against copy-paste prompting, especially when structure and revision history matter.