Measured harness ledgerPublic result
Claude Opus 5Campfire under the stars — Claude Opus 5 Max
Build a polished interactive Three.js campfire scene under a starry night with convincing fire, lighting, environmental detail, and camera motion.
Max reasoningHeadline result
- Workflow cost
- $27.45
- Wall-clock
- 39m 37.2s wall-clock
- Processed tokens
- 21.14M processed
- Record state
- partial_token_timing_artifact_build_ledger
Public summary
Claude Opus 5 Max partial_token_timing_artifact_build_ledger ledger: 39m 37.2s wall-clock, 21.14M processed, and $27.45 First-party Claude API list-price equivalent, not a subscription invoice.
Run identity and stack
- Result ID: campfire-threejs-opus-5-max
- Technical model: claude-opus-5
- Provider: Anthropic Claude Code
- Stack: Anthropic Claude Code
- Stack: Technical model/configuration: claude-opus-5
- Stack: Three.js interactive scene
- Stack: Harness v1 campfire prompt
Cost basis
- The source receipt reports that all observed cache writes used a one-hour prompt-cache TTL.
- All-5-minute cache-write alternative: $23.84.
Primary artifact integrity
- Kind: vite-threejs-campfire-source-and-production-build-entry-point
- Path: artifacts/campfire-threejs-opus-5-max/source/src/main.js
- SHA-256: cff7ff74c5b29786a94247a8b19e534eafe666c1231ae20c2178011c972aeea2
Recorded caveats
- Wall-clock is end-to-end latency, including tool execution and model reasoning; it is not model-only compute.
- Output tokens include hidden reasoning, code, and tool-call JSON, not only visible prose.
- Cache reads dominate processed-token volume but are billed at 0.1x base input, so processed tokens substantially overstate cost.
- The primary $27.45 estimate uses the source-reported one-hour cache TTL; the all-five-minute alternative is $23.84.
- The receipt's later final snapshot includes the cost of metrics and pricing work, so this ledger uses its stated canonical pre-metrics snapshot.
- The archive build passed, but no browser-runtime, local FPS, hardware, final capture, or blind-evaluation receipt was supplied.
Visible evidence gaps
- browser runtime verification
- local FPS and hardware receipt
- final capture with viewport/browser evidence
- blind-evaluation record
Public result only
This result keeps its public summary and evidence, but it does not currently have a matching Builder test with prompts, projects, Harness workflows or skills.