gpt-6POST
Debating margins on Anthropic and OpenAI economy-tier models
The poster questions whether Anthropic really runs Fable/Mythos at low margin and compares it with OpenAI's investment in economy-tier models like Astra.
Teortaxes▶️ (DeepSeek 推特🐋铁粉 2023 – ∞)
@teortaxesTexArchived 2026.09.21Originally posted 2026.09.20
Teortaxes▶️ (DeepSeek 推特🐋铁粉 2023 – ∞) @teortaxesTexOriginally posted 2026.09.20
huh It's obviously true for garbage tier models like Haiku and Sonnet, which Anthropic neglects. OpenAI cares a lot about the economy tier (thanks btw). But is this true even for Fable/Mythos though? Is Ant eating lower margin on Fable, compared to eg Astra? Hard to believe.
This post quotes another post: x.com/pequityresearch/status/2101492085039747166
x.com/teortaxestex/status/2101521728048345466Archived 2026.09.21 10:10