Skip to main content
2026/9/23 13:27 (UTC+8) · New Model · Anthropic 🚀 claude-opus-5-5 is live: Fable-class results, priced below Opus 5 The first model in Anthropic’s Claude 5.5 family, released 22 September, scores 66.4% on Terminal-Bench 4.0, ahead of Fable 5.1’s 55.8%. APIYI offers both claude-opus-5-5 and claude-opus-5-5-thinking in the default / svip / ClaudeCode groups, on both the OpenAI-compatible and native Anthropic endpoints. Priced at $4 input / $20 output / $0.20 cache reads per 1M tokens, with all four items matching the provider. Before migrating from Opus 5:
  • Thinking can’t be disabled; thinking: {"type": "disabled"} returns 400
  • tool_choice of any / tool returns 400
  • The default effort drops from high to medium; set it explicitly for coding work
📖 Full benchmarks, migration checklist and code samples: Claude Opus 5.5 Is Live
← Back to Live Updates · 📚 Monthly archive