PulseAugur
EN
LIVE 11:35:50

Anthropic's Fable 5 faces criticism over pricing and silent downgrades

A user compared Anthropic's new Fable 5 model against OpenAI's GPT 5.5 for a research project, finding Fable 5's rapid, end-to-end approach initially yielded a 30% performance boost but was unsustainable due to high API costs. The user also criticized Anthropic for silently downgrading Fable 5 usage to Opus 4.8 for certain tasks, effectively charging premium prices for lower-tier output. This contrasts with OpenAI's approach of offering free Pro subscriptions to open-source project maintainers, suggesting a divergence in strategy between the two AI labs. AI

IMPACT Highlights differing strategies in AI model access and pricing, potentially influencing user adoption and developer community engagement.

RANK_REASON User opinion piece comparing two AI models and their companies' strategies.

Read on dev.to — Anthropic tag →

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →

COVERAGE [1]

  1. dev.to — Anthropic tag TIER_1 English(EN) · guanjiawei ·

    Intelligence Is Starting to Be About Wealth

    <p>The day after Fable 5 dropped, I threw it at a research project that had been stuck for two weeks.</p> <p>Inference engine performance optimization. A 78-layer model running on hardware, baseline 10 token/s, target 100+ tps. GPT 5.5 had already sunk a lot of hours into it with…