Measured harness ledgerPublic result
Claude Fable 5Campfire under the stars — Claude Fable 5 Medium
Build a polished interactive Three.js campfire scene under a starry night with convincing fire, lighting, environmental detail, and camera motion.
Medium reasoningHeadline result
- Workflow cost
- $5.12
- Wall-clock
- 4:24.6 wall-clock
- Processed tokens
- 1.14M processed
- Record state
- partial_token_timing_and_source_ledger
Public summary
Claude Fable 5 Medium partial_token_timing_and_source_ledger ledger: 4:24.6 wall-clock, 1.14M processed, and $5.12 Anthropic first-party Claude API, standard global pricing.
Run identity and stack
- Result ID: campfire-threejs-fable-5-medium-partial
- Technical model: claude-fable-5
- Provider: Anthropic Claude Code
- Stack: Anthropic Claude Code
- Stack: Technical model/configuration: claude-fable-5
- Stack: Three.js interactive scene
- Stack: Harness v1 campfire prompt
Cost basis
- Primary cost uses 5-minute cache writes.
- All-1-hour cache-write alternative: $6.57.
Primary artifact integrity
- Kind: interactive-threejs-scene
- Path: artifacts/campfire-threejs-fable-5-medium/index.html
- SHA-256: 7d1e0181925bc21b54fa5d7b04a01444e0ad6300046d243408aed23d99455a21
Recorded caveats
- Wall-clock is end-to-end latency, including think/read time between turns, not model-only compute.
- Output tokens include hidden reasoning, code, and tool-call JSON.
- Cache-read is billed at a deep discount, so total processed tokens overstate cost.
- The cache-write TTL was not supplied; the primary total assumes 5-minute writes.
Visible evidence gaps
- final capture
- browser and hardware environment
- blind-evaluation record
Builder test available
This result is part of Builder tests. Open them for the exact prompt and any released projects, RemakeBench Harness workflows and production skills. Public proof and known evidence gaps stay visible here.
- Tech Review 001 · v1
- Kimi K3 Launch 002 · v1