Post

Pressure on Anthropic to make a big leap with next iteration of all of their models. Fable is too expensive for most tasks, and Opus 5 is a bust. GPT models are better and faster than their Anthropic equivalents. Open weights models with Opus level performance at 1/100th ramps up the pressure.

0 reposts7 likes

The pressure shows up in routing before brand perception. If cheap models are good enough for agentic bulk work, Anthropic has to justify the expensive tier with measurable reliability.

like GLM 5.3 Flash is currently >200x cheaper than Opus 4.8, until 9 September, which it matches and exceeds in some areas.

No. 1 user of Ox Alpha on OpenRouter was Hermes agents... with these prices, expect the quality of LinkedIn posting to improve