Measured harness ledgerPublic result
DeepSeek V4 FlashNeon Gunner — DeepSeek V4 Flash Max
Build a playable destructible neon run-and-gun game in the browser under the pinned Harness task contract.
Max reasoningHeadline result
- Workflow cost
- $0.20
- Wall-clock
- 5h 9m 39.389s wall-clock
- Processed tokens
- 85.64M processed
- Record state
- partial_metrics_and_validator_ledger
Public summary
DeepSeek V4 Flash Max partial_metrics_and_validator_ledger ledger: 5h 9m 39.389s wall-clock, 85.64M processed, and $0.20 OpenCode Go API-equivalent usage accounting, not an itemized subscription cash charge.
Run identity and stack
- Result ID: neon-gunner-canvas-deepseek-v4-flash-max
- Technical model: DeepSeek V4 Flash
- Provider: OpenCode Go
- Stack: OpenCode Go
- Stack: Technical model/configuration: DeepSeek V4 Flash
- Stack: Browser canvas game
- Stack: Harness v1 Neon Gunner prompt
Primary artifact integrity
- Kind: interactive-vanilla-canvas-run-and-gun
- Path: artifacts/neon-gunner-canvas-deepseek-v4-flash-max/source/index.html
- SHA-256: 5b48281672b0a30f55429f0d155a6d47488a674bdb3cd2eb947260fdfa904d13
Validation evidence
- Result: 26/26 checks recorded
- Path: artifacts/neon-gunner-canvas-deepseek-v4-flash-max/provenance/validation.public.json
- SHA-256: 251145b63f9dfc69f809dcc53770da3c4db145d71402f5b8760dd1d222ee60c3
- Validator SHA-256: fa5ec13e9ef2f5f96949283077b7c6278bf9c9f14032817c21743ab973d06090
Recorded caveats
- Wall-clock is end-to-end workflow latency including tool execution, installs, tests, browser checks, and idle gaps; it is not model-only compute.
- One OpenCode agent turn can contain many billable model requests.
- Visible output and reasoning are separately metered and both output-priced in this accounting.
- Cache reads are deeply discounted, so total processed tokens overstate cost.
- This one-shot result's token/timing receipt is PARTIAL solely because the canonical extractor needed a synthetic EOF marker on a private clone; the original database was not changed.
- The source validation log is a strong supplied completion artifact but was not independently browser-replayed while archiving.
- The $0.20 figure is an API-equivalent promotional usage value, not a marginal cash charge attributed to the OpenCode Go subscription.
Visible evidence gaps
- independent browser replay of the supplied validator run
- local FPS measurement and hardware receipt
- browser and viewport 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.