Haiku 5.5 Burns Past 100K Tokens Even on Simple Agentic Tasks, User Reports
No-Conflict4823 · reddit · 2026-10-08
A Reddit user expected Haiku 5.5's new power and pricing to beat competing models, but found it impossible to keep its token usage below 100K in any agentic use case.
Even simple, fast tasks land in the 100K–200K range, which could erode the model's headline price advantage. The poster asks whether others see the same pattern or have tricks to reduce consumption.
Related event: Haiku 5.5 draws complaints over heavy agentic token usage(3 posts)→
More from Models
- Haiku 5.5 joins AI Village, spends all day refreshing Gmail waiting for instructions — repligate · 2026-10-08
- Microsoft shows DeepSeek V4 Flash, a 284B model, running locally on Windows at 1.6-bit quantization — rohanpaul_ai · 2026-10-08
- Sentry CEO: benchmarks show nearly all code review models surface the same primary issues — zeeg · 2026-10-08
- Will AI apps start training models for their day jobs? Genspark's Gen-1 Slides on MiniMax M3 hints yes — IdekWatName · 2026-10-08
- In production comparisons keep showing Jev crushing benchmark-maxxed knock-offs — hardimanjames · 2026-10-08
- User runs GPT 6.1 Sol on max for 7 hours across 4-6 threads, still has 95% quota left — cyrus_zei · 2026-10-08