Gemini 4 Carbon checkpoint in Arena under the codename "gemini-4-argon." I asked it to build an interactive 3D Crystal Citrus with realistic textures, lighting, reflections, and animations. The output is good. Its 3D capabilities better than the Argon. Took around 20 minutes to build this.

Opus 5.5 nerfed? Today, I noticed it doesn’t feel the same as before. Its making more mistakes now. Anyone else experiencing this?

Gemini 4 Carbon Leaks: Its Not Far Away - Gemini 4 Carbon EAP has started and has been given to partners for testing. Its also being tested on LMArena, and I will do my best to get access to test this model. - Its on LMArena, so people will most likely get access to it soon. - The model is an upgrade over Argon and should fix the things Argon lacks. For now, its still bad at frontend, like Argon, but better at 3D. - It could be released at the end of this month or the beginning of November. The release isn't far away. I hope Gemini 4 Carbon is much better at frontend and coding and can match Opus 5.5. Argon was a good start, but it’s time to compete.

Gemini 4 Carbon is in development, and it's coming this month. - According to the information, Google is working on Gemini 4 Carbon, which is expected to be an upgrade over Argon. - The article also mentions that Google's models are very good at visual tasks, and I agree. When you give them an image or video to analyze, they do a really good job. - Gemini 4 Argon is still behind OpenAI and Anthropic in some areas, and I hope Google closes that gap this time. - I want Gemini 4 Carbon to be competitive with Astra and Fable, rather than just another incremental upgrade. - If the model really is coming this month, it looks like Google is picking up the pace with its releases. Carbon should reach at least Opus 5.5-level performance this time. Google needs a model that can genuinely compete with the best.

The last week of October could get crazy for AI. There are so many models that could launch soon, and I'm curious to see how many actually make it before the month ends. - Gemini 4 Carbon: Google's next upgrade over Argon. I want to see if it can finally compete with the best coding and reasoning models. - Fable 5.5: This is the one I'm watching most closely, and it can arrive as soon as this week. - GPT-6.1 Terra: coming in the next few weeks. - Qwen 4: Another release I'm keeping an eye on, with a possible launch next week. - Mistral Large 4: An open-weight release would be interesting to see. - NVIDIA Nemotron 3.5 Bolt E4B: A small on-device model expected around October 23. - A new Gemini Flash model: This could show up too. - Opus 5.6: Possible, but I'm keeping my expectations low for now for this launch. - OpenAI's three image models: These could also launch, although their current outputs haven't impressed me much. Personally, I'm most interested in Carbon and Fable 5.5. I want to see whether Google can close the gap and whether Anthropic can push its models even further. If even 70-80% of these launches happen by the end of October, it's going to be a massive month for AI. And if they don't, November could be even crazier.

Fable 5.5 looks close. - I tested the latest checkpoint, and its one shot performance is really good. - Partners already have access and say the newer checkpoint is stronger than the previous one. - With the IPO coming up, Anthropic has more reason to ship strong models quickly. - The strongest indication I'm hearing points to next week, most likely Tuesday or Wednesday. - We already know how good Opus 5.5 is, so imagine what Fable 5.5 could bring. Looks like we won't have to wait much longer.

Fable 5.5 is gonna be a really good model. I asked it to create The Last Train Home a cinematic Japanese night scene with lantern-lit streets, rain-soaked pavement, and a train passing through the city. Anthropic should launch this model ASAP.

Anthropic is preparing launch a new model, most likely Fable 5.5. - According to the information, new model is better at reasoning and intuition than their current models. - Anthropic seems to be focusing on making AI understand users more naturally, even when they don't provide complete information or explain everything perfectly. - Basically, you won't need to write long, detailed prompts just to get good results. Even with simple or half-baked prompts, the model should be able to understand your intent and figure out what needs to be done. - I think we are moving towards a phase where prompt engineering becomes less important for everyday tasks. Instead of explaining every detail, you can simply describe what you want, and the model figures out the rest. AI will feel much more natural to use.