the headlines
Developers seem to be switching from Claude to Codex
Claude Developers admitted to this
GitHub Copilot has moved to token-based pricing
Manus is not selling to Meta anymore
OpenAI’s origin details are being disputed in court
Claude Code fixed over 50+ bugs
Um, Codex has a weird system prompt
OpenAI missed key revenue targets
Codex has updated it’s UI
let’s dive in…
Claude Code Versus Codex

Developer sentiment quickly changed after the GPT-5.5 release.
The benchmarks were fairly impressive yet they didn’t raise any eyebrows, but what actually turned heads was what GPT-5.5 felt like.
Right after the release developers were already flooding onto X, Reddit and other platforms to share their positive opinions about the new GPT model.
Most considered it to be a serious upgrade from 5.4, especially when it came to speed and clarity of thought. Whilst it was 2x more expensive than it’s predeccessor, it came in at a similiar price range to Opus 4.7 and tended to consume far less tokens than the Anthropic model.
So people began to decide. Claude or Codex?
People seem to be picking the latter. Whilst switching models is fairly easy and people switch between them all the time (I still use Opus for frontend and GPT-5.5 for backend) there does seem to be a huge influx of previous Claude Code users who have made the switch to Codex.
The funny thing is, this could all be completely reverted after the next model release. Claude could release Mythos? Gemini 4 is presumably around the corner. The six-week release cadence OpenAI is now running means none of this is stable for long.
But for now, it seems that GPT-5.5 has taken a special place in developers hearts and (more importantly for OpenAI), their Codex subscription billing.
The biggest AI trial has begun

The courtroom tension ramped up fast as Elon Musk returned for day two of his legal clash with Sam Altman.
At the center of it all is OpenAI, and whether it strayed too far from its original nonprofit mission.
What initially looked like another high-profile tech dispute is starting to feel a lot more consequential.
With OpenAI now tied to an $852B valuation and a potential IPO on the horizon, the stakes stretch far beyond the courtroom.
Musk, also leading xAI and Tesla, is pushing a narrative that OpenAI’s transformation into a capped-profit powerhouse represents a fundamental betrayal of its founding principles.
Altman, on the other hand, is defending the shift as necessary to scale cutting-edge AI and stay competitive in an increasingly aggressive landscape.
Vibe coders are watching closely.
OpenAI’s APIs underpin a huge portion of the current AI ecosystem, and any disruption at the leadership level could ripple outward quickly.
There’s already quiet speculation about what happens if the court rules in Musk’s favor.
Could Altman be forced off the board?
Will that lead to changes in the API and Codex?
Everyone seems to have different opinions, but the uncertainty alone is enough to get people talking.
15x valuation jump in a 14 months

$61.5 billion to $900 billion in twelve months. Read that again.
The revenue story underneath it is what makes this valuation make sense. A $40 billion run rate, up from roughly $1 billion a year ago.
Their tactic was to aim for enterprise and power users and considering that more than 75% of their revenue comes from API billing, the tactic seems to have paid off.
50x on valuation since early 2024. Arguably the fastest value creation in tech history alongside OpenAI.
When a banker is trading his $4.8 million estate for Anthropic shares and a $5 billion cheque can't get a meeting with the CFO, you have to at least ask how much of that $900 billion is conviction and how much is just very expensive FOMO.
the timeline







