Skip to content
Home › Context Management & Reliability

Model Lineup And Pricing ⚠ verify

Context Management & Reliability

Definition

The current Claude model lineup with API IDs, context windows, max output, and per-MTok input/output pricing used to reason about cost and capability trade-offs.

Key points

⚠ verify current model IDs/prices/context sizes as of 2026-07-08 — live docs now show Sonnet 5 and Fable 5 alongside Opus 4.8, and cached study tables lag behind. Reconcile before quoting on the exam.

Live docs lineup (docs-models-pricing, 2026-07-08):

Model API ID Context Max out In $/MTok Out $/MTok
Fable 5 claude-fable-5 1M 128k $10 $50
Opus 4.8 claude-opus-4-8 1M 128k $5 $25
Sonnet 5 claude-sonnet-5 1M 128k $3 ($2 intro) $15 ($10 intro)
Haiku 4.5 claude-haiku-4-5-20251001 200k 64k $1 $5
  • Cached study tables (netlight) list: Opus 4.8 = $5/$25 (1M / 128k), Sonnet 4.6 = $3/$15, Haiku 4.5 = $1/$5. Note the study guide references Sonnet 4.6 where live docs now show Sonnet 5.
  • Sonnet 5 introductory pricing: $2 in / $10 out per MTok through Aug 31 2026; standard $3 / $15 from Sep 1 2026.
  • Model IDs from the 4.6 generation onward use a dateless but still pinned snapshot format (not evergreen).
  • effort defaults to high on Opus 4.8 (all surfaces) and Sonnet 5 (Claude API + Claude Code) — see Adaptive thinking and effort.
  • Knowledge cutoff: Fable 5 / Opus 4.8 / Sonnet 5 = Jan 2026; Haiku 4.5 = Feb 2025.

Why it matters for the exam

  • Cost/capability model-selection questions. The exam expects the right window (1M vs 200k), max output, and per-MTok rates — and awareness that IDs/prices drift.

Common gotchas

  • Cached tables may name Sonnet 4.6 while live docs show Sonnet 5 — verify which the question assumes.
  • Sonnet 5 has two price tiers straddling the Aug 31 / Sep 1 2026 cutover.
  • Haiku 4.5 is 200k, not 1M — don't assume every current model is 1M.

See also

Sources

Referenced by