Documentation
Models
If unsure where to start, use claude-sonnet-4-6 — the flagship for coding and most tasks. For maximum power use claude-opus-4-8. For creative work — claude-fable-5. All models share one key and balance.
Current models
Full pricingPrices per 1M tokens.
NEW
claude-fable-5
creative writing, storytelling, roleplay
Input price$8.00 / MTok
Output price$40.00 / MTok
SpeedThoughtfully
MAX
claude-opus-4-8
complex tasks, long context, agents
Input price$4.00 / MTok
Output price$20.00 / MTok
SpeedPowerful
FAST
claude-sonnet-4-6
daily coding, vibe coding, speed/quality balance
Input price$2.40 / MTok
Output price$12.00 / MTok
SpeedFast
LITE
claude-haiku-4-5
autocomplete, high-volume, minimal cost
Input price$0.80 / MTok
Output price$4.00 / MTok
SpeedVery fast
Which model to choose?
Choose by task, not benchmark. To switch, just change the model field.
| Task | Model | Why |
|---|---|---|
| Creative writing & storytelling | claude-fable-5 | Unique expressiveness, nuanced dialogue |
| Complex tasks & reasoning | claude-opus-4-8 | Maximum power for hard problems |
| Daily coding & vibe coding | claude-sonnet-4-6 | Best price/quality balance |
| Fast completions & autocomplete | claude-haiku-4-5 | Fastest and cheapest |
Model FAQ
Which model is best for coding?
Sonnet 4.6 is the best choice for 90% of coding tasks. For complex architecture decisions use Opus 4.8. For autocomplete and fast tasks — Haiku 4.5.
What is vibe coding and which model to use?
Vibe coding is development where you describe the task in natural language and the model writes the code. Use Sonnet 4.6 in Cursor, Claude Code, or Cline.
Which Claude model is cheapest?
Claude Haiku 4.5 — $0.80 per 1M input tokens. Ideal for autocomplete and high-volume use.
How do I switch between models?
Just change the model field: claude-fable-5, claude-opus-4-8, claude-sonnet-4-6, claude-haiku-4-5. The API key and balance are shared across all.