OpenAI’s GPT-5.6 Sol Challenges Anthropic’s Fable 5 with Cost-Effective Performance

OpenAI has launched GPT-5.6 Sol, a new flagship model that is poised to disrupt the generative AI market by offering performance nearly on par with Anthropic’s Claude Fable 5 at a substantially lower cost. This development is expected to put considerable pricing pressure on competitors, including Anthropic, and could influence AI adoption and development strategies in markets like India.

According to independent AI evaluation platform Artificial Analysis, GPT-5.6 Sol (max) achieved a score of 59 points on their Intelligence Index, just one point shy of Claude Fable 5 (max). This indicates a near-equivalent performance in aggregated benchmarks, suggesting that Sol offers a highly competitive alternative for a wide range of AI tasks.

Key facts

Metric	GPT-5.6 Sol (max)	Claude Fable 5 (max)
Intelligence Index Score	59 points	60 points
Cost per task	$1.04	Approximately $3.00+
Coding Agent Index	80 points	Lower than Sol

Agentic Coding Leadership

In a significant win for OpenAI, GPT-5.6 Sol (max) has secured the top position in Artificial Analysis’s new Coding Agent Index, scoring 80 points when operating within OpenAI’s Codex environment. This performance surpasses all other models in agentic coding tasks, a critical area for developers and enterprises relying on AI for software development. Furthermore, Sol demonstrates a reduction in output tokens for agentic coding tasks by up to 54 percent compared to models with similar performance, as noted by OpenAI CEO Sam Altman.

Cost Efficiency and Pricing Structure

One of Sol’s most compelling features is its cost-effectiveness. The model is priced at $1.04 per task, representing approximately one-third of the cost associated with Fable 5. OpenAI has also introduced smaller variants, Terra and Luna, which are 50% and 80% cheaper than Sol, respectively. This tiered pricing strategy, coupled with a new cache-write fee and a 90% discount on cache reads, aims to drive down overall operational costs for AI users. Token prices per million are set at $5/$30 (Sol), $2.50/$15 (Terra), and $1/$6 (Luna) for input/output, making advanced AI capabilities more accessible.

Market Implications and Competitive Landscape

Artificial Analysis highlights that Sol “defines a new Pareto frontier of Intelligence vs Output Tokens per Task,” indicating an optimal balance between performance and cost. This strategic move by OpenAI intensifies the ongoing price war in the AI industry. With Chinese open models, Meta’s Muse 1.1, and xAI’s Grok 4.5 already pushing down prices, OpenAI’s latest offering further squeezes competitors like Anthropic. While OpenAI’s models remain generally pricier than some of these alternatives, Sol’s performance-to-cost ratio presents a direct challenge to premium models such as Fable 5.

For Indian businesses and developers, this development means access to high-performing AI models at more competitive price points. This could accelerate the adoption of generative AI in various sectors, from software development to content creation and data analysis, fostering innovation and potentially reducing operational expenses for AI-driven initiatives. The increased competition may also encourage further advancements and cost reductions across the industry, benefiting the broader AI ecosystem in India.

Source: The Decoder, https://the-decoder.com/gpt-5-6-sol-nearly-matches-fable-5-on-aggregated-benchmarks-at-one-third-the-cost/