researchclaw-hybrid-llm-uses-haiku-not-gpt4o-mini

AutoResearchClaw’s hybrid LLM strategy uses Opus 4.6 for critical reasoning stages (hypothesis generation, paper writing, peer review — 8 stages) and Haiku 4.5 for routine stages (topic init, resource planning, export — 15 stages). Haiku 4.5 was explicitly chosen over GPT-4o-mini to keep the entire LLM stack within the Anthropic ecosystem and avoid cross-vendor API dependencies. AutoResearchClaw natively supports primary + fallback model configuration.