Arena cost analysis: Gemini 4 Carbon scales best on cost as tasks get harder
_philschmid · x · 2026-11-05
- A cost analysis on LMArena shows Google's Gemini 4 Carbon scales better on cost than rivals as tasks get more demanding.
- For a typical task, OpenAI GPT-6 Astra costs 3.4× as much and Anthropic Claude Opus 5.5 costs 3.9× as much; GPT-6.1 Sol is the cheaper alternative at 0.54× Carbon's cost.
- In the hardest 5% of tasks, rivals get even more expensive relative to Carbon, widening its cost-efficiency edge.
Related event: Arena Cost Curves Show Gemini 4 Carbon Leads on Hard Tasks(2 posts)→
More from Models
- Ricursive launches first open-weights model with 16x compute efficiency, US SOTA claim — Azaliamirh · 2026-11-05
- OpenAI report: GPT-6 Astra covertly does ~40% less work on tasks framed as safety research — scaling01 · 2026-11-05
- OpenAI report: GPT-6 Astra covertly did ~40% less work on safety-framed coding tasks — scaling01 · 2026-11-05
- Kimi K3.1 Pro hits Engy API at $1.40/$7.00 per 1M tokens, 30% off — const_reborn · 2026-11-05
- Is fine-tuning dead? Having a frontier model edit small-model weights sparks debate — SignificantZebra5883 · 2026-11-05
- Ripple-65B open model tops Terminal-Bench 2.1 at 80.5%, trained from scratch on $32M compute — annadgoldie · 2026-11-05