Gemini 3.8 Flash · JRPG boss battle presentation recording poster
The JRPG attempts split across completion, validation, and visual quality rather than producing one universal model ranking. Compare all four models
Measured harness ledgerPublic result
Gemini 3.8 Flash

JRPG boss battle — Gemini 3.8 Flash Highest · Attempt 1

Build an interactive Three.js JRPG boss battle with supplied character assets, required combat beats, animation, sound controls, effects, and a validator-ready game loop.

Highest reasoningHeadline result
Workflow cost
$1.12
Wall-clock
1530.7s wall-clock
Processed tokens
4.36M processed
Record state
complete_token_timing_artifact_and_validation_ledger
Public summary

Gemini 3.8 Flash Highest complete_token_timing_artifact_and_validation_ledger ledger: 1530.7s wall-clock, 4.36M processed, and $1.12 Provider-recorded usage estimate, not an itemized subscription cash charge.

Run identity and stack
  • Result ID: jrpg-boss-battle-threejs-gemini-3.8-flash-max-opencode-openrouter-attempt-1
  • Technical model: Gemini 3.8 Flash
  • Provider: OpenRouter
  • Client: OpenCode 1.18.23
  • Variant: max
  • Attempt: 1
  • Stack: OpenRouter
  • Stack: OpenCode 1.18.23
  • Stack: OpenCode → OpenRouter
  • Stack: Technical model/configuration: Gemini 3.8 Flash
  • Stack: Requested variant: max
  • Stack: Attempt 1
  • Stack: Three.js / Vite game
  • Stack: Supplied GLB fixture bundle
  • Stack: Harness v1 JRPG validator
Primary artifact integrity
  • Kind: manifest-verified-artifact
  • Path: artifacts/jrpg-boss-battle-threejs-gemini-3.8-flash-max-opencode-openrouter-attempt-1/source/src/main.js
  • SHA-256: cea556edbef171b1a1b8224eec2fdec06a76f242d1ea6fdcb1370f9b80763a90
Validation evidence
  • Result: pass
  • Path: artifacts/jrpg-boss-battle-threejs-gemini-3.8-flash-max-opencode-openrouter-attempt-1/verification/summary.public.json
  • SHA-256: 9be39121c46308fbb9c308579f0881f5f859bc4bbac867f6aa461be22946dcd7
Recorded caveats
  • No blind visual quality score is attached.
  • The live catalog did not expose a max variant even though OpenCode accepted and recorded max.
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.

RemakeBenchResearch console