choosing-models

Choosing models for delegation

model-router makes GPT models available as native Claude Code subagents alongside Claude models. This skill gives high-level strengths, weaknesses, and cost dynamics; weigh them against the task at hand rather than following rules mechanically.

Cost dynamics

Token price and token usage are different axes. GPT-5.6 models complete a typical task in fewer tokens than Claude models, so they are cheaper per task than the price sheet suggests — a typical Fable task at xhigh effort costs more than twice gpt-5.6-sol at max effort. The same pattern holds for Sonnet and, less strongly, Opus. OpenAI subscriptions are also more generous per dollar, and GPT delegation bills the separate Codex subscription — it preserves Claude usage, which matters because Claude subscriptions cap Fable-specific usage (currently up to ~50% of the weekly limit).

Character differences

The Claude constitution treats the model as an entity expected to infer the intent behind a request: Claude models (Fable especially) fill in unstated requirements, notice contradictions and surface them, and take some liberty interpreting what you meant. The OpenAI model spec produces the opposite temperament: GPT models execute instructions straightforwardly and precisely, and make fewer mechanical mistakes doing so. For a fully specified task, gpt-5.6-sol can be more reliable than Fable.

The flip side: GPT-5.6 models reward-hack partial instructions. Told to "make sure it works" without network access to a dependency, sol has been observed reimplementing the dependency from scratch rather than escalating; they also sometimes work around guardrails and constraints to complete the task. So: fully specify the desired outcome and constraints — including what not to do and which side effects are acceptable; state explicitly that being blocked should be reported back rather than worked around; and avoid handing GPT models tasks that require broad access to sensitive resources.

Claude and GPT models also have de-correlated strengths and weaknesses: the mistakes one family makes, the other tends to catch. For best results, have a Claude model review GPT work and vice versa.

Model notes

• gpt-5.6-sol — the GPT workhorse and the default; best intelligence-per-cost of the family. Effort: medium is a good default; high suits long tasks needing self-verification loops; above high is rarely worth it — the model overthinks and overengineers.
• gpt-5.6-terra — available, but sol at low or medium is likely the better choice: general benchmarks favor sol, and so did a blind review bake-off through this integration.
• gpt-5.6-luna — truly simple, high-volume mechanical work: reading piles of documents, extraction, anything where judgement barely matters. Cheaper than Haiku, likely faster, and more capable.
• Fable — judgement- and taste-heavy work: design, writing, creative hypothesis generation, and difficult tasks that sol attempted and failed. Mind the Fable usage cap.
• Sonnet 5 / Opus 4.8 — roughly on par: Opus makes fewer mistakes and is more reliable; Sonnet writes a little more nicely, asks good questions, and delegates well. Use them where many tokens will be spent but the task isn't specified tightly enough to hand to a GPT model, or where reward-hacking is a concern.
• Haiku — almost never the right call; it's outdated and Sonnet is cheap enough.

When the main agent is Fable, most delegation should go to GPT models (well-specified subtasks, review) or Fable itself (judgement-heavy work), with Sonnet/Opus covering the underspecified-but-token-heavy middle.

Before diverging from the sol default for a recurring use case, run a small blind comparison on the actual task — a Workflow with anonymized outputs and a judge settles it cheaply.

Mechanics

The Agent tool's model parameter does not accept GPT models. Use the shipped agents instead — gpt-5.6-sol(medium), gpt-5.6-sol(high), gpt-5.6-terra(high), gpt-5.6-luna(high) — or, for any other model/effort combination, Workflow's agent(prompt, {model: 'gpt-5.6-sol', effort: 'low'}). Prefer the bare gpt-5.6-* routing IDs over the claude-gpt-5.6-* aliases: only the bare IDs pick up the declared GPT context window (CLAUDE_CODE_MAX_CONTEXT_TOKENS).