GPT-6.1 Sol sets a new cost-efficiency frontier on Intelligence Index

• Artificial Analysis measures GPT-6.1 Sol at max effort at $0.72 per Intelligence Index task, 31% below GPT-6 Sol ($1.04) and 64% below GPT-5.6 Sol ($1.99).
• Every effort level of GPT-6.1 Sol sits on the cost-efficiency Pareto frontier in Artificial Analysis's testing: for a given intelligence score, it found no cheaper model.
• OpenAI fixed an image-encoding bug in GPT-6 Luna and Sol. Artificial Analysis says Luna (max) gained a point overall, with vision-heavy evals up most (MMMU-Pro +4.1, GDP.pdf +2.4), while Sol barely changed.

• eval — A repeatable test for AI quality — a set of tasks plus scoring — used the way software teams use test suites, because model output is too variable to judge by eyeballing.
• AI agent — An AI system that doesn't just answer once but works toward a goal in a loop — taking actions, reading the results, and deciding what to do next.

GPT-6.1 Sol costs $0.72 per Intelligence Index task at max effort versus $3.26 for GPT-6 Astra, and no cheaper model matches its intelligence at any effort level. Useful for cost-driven model selection in agent workloads.

• postSafety blocks hit defensive cyber work: CyberGym results and costs
• postSolar Mini 4: cheap tokens, expensive tasks, weak agentic coding
• postArtificial Analysis launches Cyber Index for AI cyber defense evaluation
• postGPT-6.1 Sol reaches near-Astra intelligence at a quarter of the cost per task

• postGPT-6 Sol and Luna roughly halve OpenAI's cost-per-intelligenceArtificial Analysis
• postArtificial Analysis Benchmarks GPT-6 Sol and Luna's Cost CutsArtificialAnlys
• postGPT-6.1 Sol reaches near-Astra intelligence at a quarter of the cost per taskArtificial Analysis

Checking sign-in…

Loading comments…