haiku-45-for-routine-pipeline-stages-not-gpt4o-mini

When routing LLM calls across pipeline stages, prefer Haiku 4.5 over GPT-4o-mini or Gemini Flash for routine/low-reasoning stages to stay within the Anthropic ecosystem. This avoids cross-provider API key management, aligns with the existing Max subscription, and keeps model governance centralized. Opus 4.6 handles the 8 stages requiring deep reasoning; Haiku 4.5 handles the 15 routine stages.

Related

• haiku-45-preferred-over-gpt4o-mini-for-routine-stages-anthro
• autoresearchclaw-hybrid-llm-routing-opus-haiku
• researchclaw-hybrid-llm-opus-haiku-stage-split
• autoresearchclaw-hybrid-llm-opus-for-critical-stages-only
• hybrid-opus-haiku-for-researchclaw-stay-anthropic