SpaceXAI has added SWE Marathon benchmark to Grok 4.5 blog Grok 4.5 ranks #1 in SWE Marathon

@techdevnotes Space XAI be like

@techdevnotes grok won a benchmark that appeared the same day as grok. bold strategy

Your first Gopuff order gets $0 delivery, plus 100s of deals once you're in. 5,000+ products, no markups, delivered in as fast as 15 minutes.

@techdevnotes Grok is a great success.

@techdevnotes no cap, that’s wild. grok just went full alpha

Context: this is the one benchmark of five in the Grok 4.5 blog where it leads. Fable max tops the other four. The real Grok 4.5 story is economics: $2 per M input, 80 TPS, and reportedly ~4x fewer output tokens per task than Opus 4.8. Winning on cost per solved task, not raw score

@techdevnotes The rate of AI progress is moving at a dangerous speed. 🤔 Maybe we should slow down things a bit?

@techdevnotes arena.ai/leaderboard/

@techdevnotes Rankings change fast, but the competition benefits everyone.

@techdevnotes @grok help me to understand

@techdevnotes vip pro max

Shop luxury fashion finds at Amazon 👜

@techdevnotes Grok 4.5 dominates SWE Marathon!

@techdevnotes Seems good

@techdevnotes Nice to see Grok 4.5 killing it in SWE Marathon, that's a big win for AI research

@techdevnotes Impressive result from @xAI

@techdevnotes 지표대로 이뤄내실 수 있는거죠? 믿어 볼게요.

@techdevnotes This is wow

@techdevnotes #1 for a reason.

@techdevnotes @grok que hace este benchmark?

@techdevnotes x.com/i/status/20751…