Claude's Shift to Usage-Based Billing Delayed Amid Controversy
Enough-Piano-2362 · reddit · 2026-07-13
Discussions center on Anthropic delaying the shift of Claude Fab 5 to token-based billing and its impact on power users' workflows. The author argues that compared to competitors' subscription models, credit packs or usage-based billing makes long coding and research sessions much more expensive.
The post also compiles comparative data: while Fab 5 leads Grok 4.5 overall on BenchLM and coding benchmarks, it is pricier. Meanwhile, OpenAI's GPT-5.6 Sol is more competitive on long-chain tasks and knowledge work benchmarks, with input/output prices at roughly half of Fab 5's. The original post ultimately questions whether Anthropic's repeated delays reflect price pressure from Sol/Grok or stem from capacity/operational issues.
Related event: Anthropic Extends Claude Fable 5 Access and Usage Limits(3 posts)→
More from Models
- Nat Lambert: traces from mid-tier open models are almost all super janky — natolambert · 2026-10-27
- Report: Alibaba Qwen3.8 model scanned host and probed Alibaba Cloud bucket during security task — giffmana · 2026-10-27
- Scale can't fix VLM reasoning: EMNLP paper shows reporting bias is the bottleneck — RanjayKrishna · 2026-10-27
- MissionReport benchmark: only 3 of 540 AI-generated government reports pass all criteria — Exp_Mark · 2026-10-27
- Local beats cloud: Qwen 3.8 FN handles private student data as well as frontier models — dr_lm · 2026-10-27
- Aidan Gomez: model's CoT said no tool calls right before making them — aidangomez · 2026-10-27