Model Recommendations
Complexity score (0-100)
ctok computes a complexity score from the input before picking a model:
| Signal | Max weight |
|---|---|
Task type base (bug-fix → architecture) | 12-60 |
| Input volume (log-scaled) | 35 |
| File breadth (number of attached files) | 20 |
| Content heterogeneity (≥3 kinds) | 6 |
Keyword signals (security, distributed, migrate, …) | ±12 each |
| Large expected output | 8 |
Bands: simple < 30 ≤ normal < 65 ≤ deep
Model selection
| Band | Recommended model | Why |
|---|---|---|
| Simple | haiku-4-5 | Narrow edits, cosmetic, docs - Haiku is fast and cheap. |
| Normal | sonnet-4-6 | The default workhorse - strong reasoning, good price. |
| Deep | opus-4-7 | Architecture, hairy bugs, cross-cutting refactors. Worth the premium. |
When input exceeds 150k tokens, the recommender nudges toward Sonnet or Opus regardless of complexity band (long-context tasks benefit from stronger models).
Effort levels
Effort starts from the complexity band and is adjusted by task type:
| Task type | Adjustment |
|---|---|
architecture | → xhigh |
debugging | → high |
feature | +1 level from band |
review | → low or medium |
documentation | → low |
bug-fix | no override |
refactor | no override |
general | no override |
Levels: low → medium → high → xhigh
Pricing reference
| Model | Input | Output |
|---|---|---|
haiku-4-5 | $0.80/M | $4.00/M |
sonnet-4-6 | $3.00/M | $15.00/M |
opus-4-7 | $15.00/M | $75.00/M |
Anthropic list rates as of May 2026. ctok shows estimates only - verify at anthropic.com/pricing.