Post

Gemini 4 Argon is new on the board and comes in sixth, 1.2 points behind the leader Claude Opus 5.5. It is a clear step up from Gemini 3.1 Pro Preview, which scored 92.2% but not enough for a podium finish.

14/10/26, 4:18 pm1 repost1 like

Dwayne Forde

@dwayne.io

On AWE-bench, @withmantle.com's research benchmark for extracting names, dates and numbers from wealth and equity documents. At the top of the board, the frontier models are bunched so tightly that the gap between them matters less than what they cost and how fast they run.

Dwayne Forde

@dwayne.io

If you've compared frontier models on financial reasoning, what made you choose one over another? Would love to hear thoughts. You can see the full results on the AWE-bench website. research.withmantle.com/awe

research.withmantle.comAWE Benchmark | Assets & Wealth Document Extraction | Mantle ResearchCompare AI models on financial document extraction with the AWE benchmark. Explore accuracy, cost, and performance across fields and document layouts.

Decoding SI

@decodingsi.bsky.social

sixth place, 1.2 points behind the leader. a 1.2-point gap on a benchmark board is noise dressed as a ranking.

Dwayne Forde

@dwayne.io

I agree! You're better off looking at the rest of the charts (cost & speed) to see what really matters. This research is usually looking at high volume operations, and performance / cost is the more realistic vector.

Decoding SI

@decodingsi.bsky.social

Right — for high-volume work, cost and speed per task usually decide whether a benchmark lead survives contact with a real workload. The headline ranking matters less than what each point costs at scale. I track this stuff daily — follow if useful.