The four models
| Model ID | Context | Input (credits) | Output (credits) | List price (input / output) |
|---|---|---|---|---|
claude-opus-5-5 | 1M | 400 | 2,000 | $4 / $20 |
claude-fable-5-1 | 1M | 1,000 | 5,000 | $10 / $50 |
claude-sonnet-5-5 | 1M | 200 | 1,000 | $2 / $10 |
claude-haiku-5-5 | 1M | 10 | 50 | $0.10 / $0.50 |
Prices are per 1M tokens. 1 credit = $0.01 of usage at list price, so the credit price is the USD list price times 100, and 100 credits cost you $0.15 (15% of list price). All four models have a 1M context window and a 128K output limit. Haiku 5.5 switches to Anthropic's long-context price when a prompt exceeds 100K tokens: 50 credits input and 250 credits output per 1M. For a side-by-side comparison, see the model overview; full prices are on the pricing page.
How to switch
- In a session: type
/model opus,/model fable,/model sonnetor/model haikuto switch immediately. - At startup:
claude --model sonnetapplies to that session only.
# at startup
claude --model opus
claude --model fable
claude --model sonnet
claude --model haiku
# inside a session
/model opus
/model fable
/model sonnet
/model haikuAlias mapping
| Claude Code alias | Model | Best for |
|---|---|---|
opus (also the default) | claude-opus-5-5 | Flagship workhorse: complex coding, architecture, long tasks and hard bugs |
fable | claude-fable-5-1 | Anthropic's most capable model: the hardest reasoning and long-horizon agent work |
sonnet | claude-sonnet-5-5 | Balanced partner: subagents, bulk edits and everyday development |
haiku | claude-haiku-5-5 | Fast and light: small tasks, session titles and quick questions |
No model-mapping configuration is needed. Setup uses just two environment variables, ANTHROPIC_BASE_URL and ANTHROPIC_AUTH_TOKEN; see Claude Code setup. For tips on splitting work across the four models, see how to choose a model.
Check the current model
Type /model in a session to see the current selection. The console usage log also shows the model each request used.