Anthropic Data Shows Companies Won't Pay for the Best Models
sudoraohacker · x · 2026-08-27
Anthropic's revenue data reveals that only 11% comes from their strongest model, Fable, while the majority goes to Opus 4.8/5. This indicates businesses want strong models but don't find the top tier worth the premium. Opus 4.8/5 performs similarly to Chinese open-source models like Kimi K3 and GLM-5.3, which are cheaper. This suggests strong underlying demand for Chinese open-source models if US businesses can overcome regulatory concerns.
Related event: Enterprises Shun Top-Tier Models as Open-Source and Chinese Rivals Gain(3 posts)→
More from Venture
- Ex-NVIDIA AI Chief Sanja Fidler raises $90M seed for Veeda AI — ZGojcic · 2026-08-27
- Google Cloud adds pay-as-you-go pricing for Gemini Enterprise — rseroter · 2026-08-27
- Grafana hits $600M ARR, up 50%, driven by AI deployment — brucemacv · 2026-08-27
- bidbook.lol: a leaderboard you pay to climb — $41 takes the #1 spot — ThePeterMick · 2026-08-27
- Instead of a $144 Grammarly renewal, he built his own with Replit and MCP in a day — amasad · 2026-08-27
- DeepSeek launches ultra-cheap coding model; OpenAI slashes Luna price by 80% — nikvassev · 2026-08-27