Pricing date
2026-09-19 table
The estimate table uses official Claude API pricing checked on September 19, 2026, covering Fable 5.1, Opus 5, Sonnet 5, Haiku 4.5, and legacy rows.
Free browser tool
Claude Code Token Usage Analyzer
Estimate Claude Code transcript token usage and cost by turn with a dated model pricing table and CSV export.
Boundary
Not a bill
The analyzer is for local review and anomaly detection, not authoritative billing or chargeback.
Usage turns0
Total tokens0
Estimated cost$0.00
Unpriced turns0
Select a transcript to calculate usage and estimated cost.
| Line | Model | Input | Cache write | Cache read | Output | Estimate |
|---|---|---|---|---|---|---|
| No transcript loaded. | ||||||
Pricing table
Rows cover the current lineup (Fable 5.1, Opus 5, Sonnet 5, Haiku 4.5) plus legacy rows for older transcripts. Cache reads price at 2.5% of base input on Fable 5.1 and Mythos 5.1 and at 10% on the other rows, so a cache-heavy Fable 5.1 session can cost far less than raw token counts suggest.
FAQ
Which token fields does the analyzer use?
It uses numeric usage fields such as input_tokens, cache_creation_input_tokens, cache_read_input_tokens, and output_tokens when present on transcript records.
Can this estimate differ from my Claude bill?
Yes. Pricing date, product plan, provider, region, caching behavior, batch discounts, and official billing reconciliation can all differ from a local estimate.
Which models does the pricing table cover?
Claude Fable 5.1, Claude Opus 5, Claude Sonnet 5, and Claude Haiku 4.5 as the current lineup, plus legacy rows for Fable 5, Opus 4.8 through 4.5, and Sonnet 4.6 and 4.5 so older transcripts still price correctly. Auto-detection reads the model field on each turn and falls back to an unpriced row when it does not recognize the identifier.
Related evidence
Free tool
Log Viewer
Drop a Claude Code JSONL transcript and turn it into a readable local timeline of messages, tool calls, tool results, and token fields.
Free tool
Audit Checklist
Run an interactive agent-session audit checklist and export a concise Markdown note for PRs, incidents, and handoffs.
Bibliography
Sources
Annotated source list for Claude Code session logs, Agent SDK session storage, hooks, telemetry, and general log-management guidance.
Foundation
Session Logs
A source-grounded explanation of Claude Code and Agent SDK session logs, what they contain, and what they do not prove.
Field guide
Transcript Anatomy
How to read Claude JSONL transcript artifacts, hook transcript paths, subagent transcript paths, and stream-json output without mixing them up.