Is Haiku 5.5 good enough to be your main coding agent, not just the cheap one? I ran a small controlled test
I tested Claude Haiku 5.5 at low, medium and high effort against Sonnet 5.5 (medium), GPT-6 Luna (high) and GPT-6.1 Sol (medium).

A compact Claude model for fast, high-volume work
The overview
Around the internet
Public experiences, questions, and discussions.
Inside the toolkit
Anthropic positions Haiku 5.5 for high-volume work where speed and cost efficiency matter.
Haiku 5.5 is the first Haiku with effort controls, allowing teams to tune cost against intelligence for each task.
Anthropic lists classification, summarization, text generation, chat, voice agents, live support, and in-app assistants.
The page describes Haiku 5.5 for defined subagent work, repetitive computer use, focused coding, and multi-step tool use.
Anthropic lists Claude and cloud-provider access for Haiku 5.5.
Where it fits
Keep exploring