Anthropic says Sonnet 5.5 at low/medium effort beats Sonnet 5's best scores at ~1/10 the cost
airesearch12 · x · 2026-09-29
Anthropic's official account announced that Sonnet 5.5, running at Low or Medium effort, beats Sonnet 5's best benchmark scores on several benchmarks while costing about a tenth as much — the company's core pitch for the new mid-tier model's price-performance.
More from Models
- Agent Arena Launches Agent Leaderboard Ranking 46 Models on 2M+ Real-World Sessions — arena · 2026-09-29
- Opus 4.5 Mention Sparks Buzz on X — repligate · 2026-09-29
- Sonnet 5.5 Tops Every's Readability Scores, Feels Like Opus 5.5 and Fast — every · 2026-09-29
- Qwen-Next 3.8 Reportedly Matches Sonnet 5.5 on Coding as Local Models Close the Gap — LegacyRemaster · 2026-09-29
- Sonnet 5.5 xhigh hits Fable 5.1-level performance at a third of the cost — theimposingshadow · 2026-09-29
- Sonnet 5.5 pushes the last OpenAI models off category-split LLM Pareto frontiers — DecidingToBeTheSame · 2026-09-29