SpaceXAI Just Entered the AI Race
You will know what changed when xAI became SpaceX AI and released Grok 4.5 on GPT 5.6 launch day, how it compares on benchmarks and cost, and how to test the three leading models yourself.
What you need
A Grok account (grok.com)
A ChatGPT account for GPT 5.6
Access to Fable 5
One real task from your business to test with
Step by step
See what changed on launch day
Note the news: on the same day GPT 5.6 was released, xAI became SpaceX AI. After SpaceX bought Cursor, it released Grok 4.5, a model I call nearly as good as Opus 4.7 and trained in very little time.
Understand why the combination matters
Look at who is behind the model: SpaceX resources, the xAI team and the Cursor teams working together. My point is that this combination builds state-of-the-art models at roughly one tenth of the cost of the others.
Check the Opus-class claim
Take Elon's "Opus-class" label at the right level: Grok 4.5 is close to Opus 4.7 but not yet at Opus 4.8. What stands out is the short training time and the cost.
Compare benchmarks and cost per task
Read the comparison: Grok 4.5 does not beat Fable 5, but on SWE-bench it comes out ahead of GPT 5.5. On cost per task, Fable 5 takes about $2.75, and Grok 4.5 is nearly half that.
Treat it as a three-player race
Stop thinking in terms of two leaders. Your options are now Fable 5, GPT 5.6 and Grok 4.5. Grok 4.5 is a serious contender because of its cost and how fast SpaceX AI is putting resources into new models.
Read early sentiment with a filter
Scan reactions on X. Most people say positive things about Grok 4.5, but some of those posts may be paid. Use the sentiment as a signal, not proof.
Try all three models yourself
Run the same task in Fable 5, GPT 5.6 and Grok 4.5 and compare the results for your own work. I see this as a category shift, not a single model launch.
Grok (xAI)https://grok.com
ChatGPT / OpenAIhttps://openai.com
Watch for Google's answer
Keep an eye on Google: when this video was recorded, it had no new Gemini model yet. Whether Google holds the third spot will become clear over the next couple of months.
Google Geminihttps://gemini.google.com
Watch out for
Grok 4.5 is close to Opus 4.7 but still not as good as Opus 4.8, so do not expect it to top every chart.
Some of the positive posts about Grok 4.5 on X may be paid, so do not rely on social sentiment alone.
Things move fast: as I say in the video, nobody knows what the next month will bring, so check results again before you commit to one model.
