NewsBite analysis · AI, Business
The Decoder
GPT-5.6 Sol nearly matches Fable 5 on aggregated benchmarks at one-third the cost
OpenAI's newly released GPT-5.6 Sol model demonstrates competitive performance, achieving scores almost equal to those of the Fable 5 model while significantly reducing costs. Independent evaluations show that Sol excels in coding tasks and has introduced innovative pricing structures, such as cache-write fees, which further enhance cost-efficiency. This advancement signals a new trend in AI models where performance gains are coupled with reduced pricing, challenging existing competitors and reshaping market dynamics.
MonitorConfidence: HighShort term horizon
Published 9 Jul 2026, 20:10 UTC1 min read
Visit original article ↗Lead image — The Decoder
Summary
TL;DR
- GPT-5.6 Sol scored 59 points on the Intelligence Index, just behind Fable 5.
- Dominates the Coding Agent Index with a score of 80 points.
- Costs $1.04 per task, one-third the price of Fable 5.
- Introduces cache-write fees to decrease pricing further.
- Smaller variants, Terra and Luna, are priced significantly lower than Sol.
- Pricing pressure from competitors could lead to a market-wide price cut.
Interesting facts
- Sol burns up to 54% fewer output tokens on similar tasks compared to competitors.
- New pricing strategies could challenge the pricing structures of established AI models.