GPT-5.6 Sol, Grok 4.5, Muse Spark 1.1 and Fable 5: The Week the AI Race Exploded
Published on
GPT-5.6 Sol, Grok 4.5, Muse Spark 1.1 and Fable 5 available until the 12th, all of that in less than 48 hours. This week, the AI model race EXPLODED!

The biggest launch of the week (literally today, July 9) was OpenAI’s GPT-5.6, which finally reached the general public after a few weeks restricted to select partners, at the request of the American government (a consequence of the Fable drama). The model comes in three versions: Sol, Terra and Luna (copying the conceptual style of Anthropic’s Opus/Sonnet/Haiku), with Sol Ultra being the most powerful version, with a mode similar to Claude’s “Ultracode with dynamic workflows”, which delegates parts of the tasks to specialized submodels.
Sam Altman said Sol is 54% more token-efficient on agentic coding tasks compared to GPT-5.5, which is excellent news for anyone paying the API bill every month. All the early testers I saw on social media loved it, even comparing it to Fable on some tasks.
SpaceXAI (yes, the merger of Elon’s xAI with SpaceX) also took advantage of the week to launch Grok 4.5. Musk said it’s on the same level as Claude Opus, but that this Grok 4.5 is faster, more token-efficient and cheaper. What the early testers said is that it reaches the level of Opus 4.6 or 4.7, which is already enough for a lot of people, and since it’s way cheaper than Opus 4.8 the advantage is even better. I heard a lot of praise.
And Meta showed up too. Besides a “leak” saying their next model (codename “Watermelon”) had reached GPT-5.5 in performance, today they launched the Muse Spark 1.1 model, an update focused on code and agentic tasks that also gets very close to Opus 4.8 on the benchmarks (and even surpasses it on some). API access is limited for now.
And what about Google DeepMind’s new Gemini? They promised it for last month, pushed it to this one (probably because they weren’t happy with the results). Apparently it’s coming soon. The suspense remains.
And the best part for us is that Anthropic responded to the launch of GPT-5.6 Sol and extended access via account (and not via API) to Claude Fable 5 until July 12. And best of all: it reset everyone’s 5h and weekly limits! To my joy, because I’m using Fable on Max effort and was already at 86%. 😅
It’s a VERY busy week, and Marcus Mendes and I are going to analyze and break all of this down (and much more) in tomorrow’s episode of IA Sob Controle.
Let them cook!