Measured harness ledgerPublic result
Claude Sonnet 5Campfire under the stars — Claude Sonnet 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
- $19.22
- Wall-clock
- 49m 25.0s wall-clock
- Processed tokens
- 63.92M processed
- Record state
- partial_token_timing_and_artifact_ledger
Public summary
Claude Sonnet 5 Max partial_token_timing_and_artifact_ledger ledger: 49m 25.0s wall-clock, 63.92M processed, and $19.22 First-party Claude API list-price equivalent, not a subscription invoice.
Run identity and stack
- Result ID: campfire-threejs-sonnet-5-max
- Technical model: claude-sonnet-5
- Provider: Anthropic Claude Code
- Stack: Anthropic Claude Code
- Stack: Technical model/configuration: claude-sonnet-5
- Stack: Three.js interactive scene
- Stack: Harness v1 campfire prompt
Cost basis
- The supplied receipt records introductory Claude Sonnet 5 prices effective through 2026-08-31; the run date is inside that period.
- The source receipt states that the session used a one-hour prompt-cache TTL.
Primary artifact integrity
- Kind: self-contained-interactive-threejs-campfire-html
- Path: artifacts/campfire-threejs-sonnet-5-max/campfire_starry_night.html
- SHA-256: 7f091584443a0053b10a9dcc673c8d5638f87a5e5ccecf453d93c716e88709cd
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, generated code, and tool-call JSON, not only visible prose.
- Cache reads dominate processed-token volume but are billed at a deep discount, so processed tokens substantially overstate cost.
- The primary wall-clock interval includes one nested subagent. Its tokens and turns are included, but its duration is not added a second time.
- The source receipt's generic newest-transcript heuristic was corrected to avoid an unrelated concurrently active session; the correction is documented in the sanitized sidecar.
- The supplied artifact was hash-pinned without a fresh browser run in this environment. No local FPS, browser/hardware environment, final capture, or blind evaluation was supplied.
Visible evidence gaps
- browser runtime verification
- browser and hardware environment
- local FPS measurement receipt
- final capture
- 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.