Blog
4 min read

How to Change the Model in Claude Code (and Which One to Pick)

Switch models in Claude Code with /model, the --model flag, an environment variable or settings.json. What the aliases mean — opus, sonnet, haiku, fable, opusplan, [1m] — which is the default, and how to choose for cost and quality.

Claude Code doesn't lock you to one model. You can switch mid-conversation, start a session on a specific model, or set a default for every session. Choosing well is one of the easiest ways to make Claude Code faster or cheaper without making it worse.

Model names and defaults below match Anthropic's model-configuration docs as of October 2026. New models arrive often; the aliases keep working when they do.

The four ways to change model

From highest to lowest priority:

1. During a session: /model

Type /model to open a picker. Press Enter to switch and save it as your default, or s to switch for this session only. Or skip the picker:

/model sonnet
/model opus

2. When starting: --model

claude --model sonnet

3. An environment variable

export ANTHROPIC_MODEL=sonnet

4. Your settings file

In ~/.claude/settings.json (or a project's .claude/settings.json):

{
  "model": "sonnet"
}

See Claude Code settings explained for how user, project and local settings combine.

To check which model you're on, run /status, or add it to your status line.

What the aliases mean

You'll rarely type full model IDs. Aliases always point at the latest model of each kind:

Alias What it means
default Removes any override; uses your account's default
opus The latest Opus — strong reasoning for complex work
sonnet The latest Sonnet — fast and capable for everyday coding
haiku The fast, cheap model for simple tasks
fable Anthropic's most capable model, for the hardest, longest tasks (where available)
best Fable where available, otherwise Opus
opusplan Opus while planning, Sonnet while carrying out the plan
opus[1m], sonnet[1m] A 1-million-token context window on models that need the suffix

On the Anthropic API and on Pro, Max, Team and Enterprise plans, the default is currently Opus 5.5. The newest Sonnet, Opus and Fable models already have a 1-million-token context window without the [1m] suffix; the suffix matters for some older models.

Which model should you use?

A simple rule from Anthropic's own guidance: Sonnet handles most coding tasks well and costs less than Opus. Reserve Opus for complex architecture and multi-step reasoning.

In practice:

  • Everyday edits, small features, bug fixes with a clear cause: Sonnet.
  • Designing something new, a gnarly bug, a refactor across many files: Opus (or opusplan to plan with Opus and execute with Sonnet).
  • Very long, open-ended or high-stakes work: Fable, if your plan includes it.
  • Simple, repetitive jobs, especially in subagents: Haiku. You can set model: haiku in a subagent's configuration.

Effort is the other dial

Each model also has an effort level — how much it thinks before answering. Lower effort means fewer tokens and faster replies on easy tasks. Change it with /effort or inside /model. On the newest models, thinking can't be turned off entirely, but lowering effort has a similar effect.

Models and usage limits

On a subscription, there are two kinds of limit:

  • Session and weekly limits are shared across all models. Switching model won't get you around them.
  • Model-specific limits — a message like "You've hit your Opus limit" — can be worked around by switching to another model family with /model.

Using Sonnet for routine work makes your plan's limits last noticeably longer. On an API key, it lowers the bill directly. See how to keep Claude Code costs down for more.

The summary

  • /model switches mid-session; --model, ANTHROPIC_MODEL and "model" in settings set it up front.
  • Use aliases (sonnet, opus, haiku, fable, opusplan) rather than full IDs.
  • Default is Opus 5.5 today; Sonnet is the cost-effective choice for most coding.
  • Lower effort for simple tasks; switch families if you hit a model-specific limit.

EasySpawn hosts Claude Code on a persistent server and lets you bring your own Claude subscription or API key — so your model choice, settings and usage stay entirely yours. See how it works or join the waitlist.

Related: Claude Pro vs Max vs API Key · What Are Tokens in AI? · Claude Code /compact vs /clear · What Is an LLM?

Keep reading