November 5, 2026

Anthropic’s Model 4 becomes highest-scoring AI model, clears 50% on all benchmarks

Anthropic has completed training its new internal model, referred to as Model 4, which has achieved the highest scores in METR’s analysis, becoming the first model to surpass 50% on every benchmark evaluated. Additionally, METR suggests that Anthropic has likely exceeded the 2x acceleration threshold under its Responsible Scaling Policy, potentially operating at up to 3x, with an estimated 80% of the current acceleration in AI progress attributed to advancements in the field itself as of October 2026.

METR: METR is an organization specializing in the evaluation of frontier AI models for capabilities, risks, and scaling trends. In this development, METR analyzed Anthropic’s latest model and concluded it represents a notable acceleration under Anthropic’s Responsible Scaling Policy.
Anthropic: Anthropic is an AI company dedicated to building advanced, safe artificial intelligence systems. In the current news, the company recently completed training on a new internal model that METR has evaluated as Model 4 and identified as the top performer among those tested.
Andrew Curran: Andrew Curran is an observer and commentator focused on AI developments and evaluations. He publicly shared METR’s assessment of Anthropic’s model performance and acceleration metrics in relation to responsible scaling thresholds.

Scaling Acceleration: METR assesses that Anthropic has reached or exceeded the 2x acceleration threshold in its Responsible Scaling Policy and may now be operating at up to 3x.
Benchmark Performance: The new Anthropic model achieved the highest scores in METR’s analysis and became the first to exceed 50% on every benchmark evaluated.
AI Contribution Estimate: METR estimates that approximately 80% of the observed acceleration in AI progress as of October 2026 stems from advancements in AI itself.

AI tools built by Emerald Force

Built and supported by Emerald Force.

You might also like