Measured harness ledgerPublic result
Qwen3.8 Max Preview

Interactive Three.js voxel map viewer — Qwen3.8 Max Preview Unreported · Attempt 1

Build a Three.js voxel map viewer with selectable Jungle Temple, Oasis Outpost, and Shrine Village maps, camera reset behavior, and OrbitControls verification.

Unreported reasoningHeadline result
Workflow cost
$13.61
Wall-clock
45m 25.4s wall-clock
Processed tokens
8.71M processed
Record state
published_artifact_runtime_ledger
Public summary

Qwen3.8 Max Preview Unreported published_artifact_runtime_ledger ledger: 45m 25.4s wall-clock, 8.71M processed, and $13.61 Qualified third-party NanoGPT API-list-price scenario for the measured token mix; not an Alibaba Token Plan cash charge.

Run identity and stack
  • Result ID: interactive-voxel-map-viewer-qwen3.8-max-preview-attempt-1-isolated-vision
  • Technical model: qwen3.8-max-preview
  • Provider: Alibaba Cloud Model Studio
  • Client: Qwen Code 0.20.0
  • Attempt: 1
  • Stack: Alibaba Cloud Model Studio
  • Stack: Qwen Code 0.20.0
  • Stack: Technical model/configuration: qwen3.8-max-preview
  • Stack: Attempt 1
  • Stack: Three.js voxel viewer
  • Stack: OrbitControls
  • Stack: Harness v1 voxel-viewer prompt
Cost basis
  • No first-party Qwen3.8 cache rate was established, so cached input uses the published third-party input rate. This is not Alibaba pricing, provider cost, or an itemized subscription charge.
  • Qualified API-list-price equivalent.
Primary artifact integrity
  • Kind: interactive-threejs-voxel-map-viewer
  • Path: artifacts/interactive-voxel-map-viewer-qwen3.8-max-preview-attempt-1-isolated-vision/source/index.html
  • SHA-256: f98aa7bd1b1f23c903609cda2085c717252418c97d5dd738348831f31a159ba5
Validation evidence
  • Result: PASS
Recorded caveats
  • Wall-clock is end-to-end workflow latency, not model-only compute time.
  • Output tokens include Qwen-accounted thoughts/reasoning, visible prose/code, and tool-related output; 109,749 thought tokens are a subset of 158,946 output tokens.
  • The actual captures are 1600×913; 1600×1000 was the requested Chrome window size.
  • The 120.08 FPS sample is a single Apple M1 Pro / ANGLE Metal browser measurement, not a standardized cross-machine comparison.
  • The app requires network access to load Three.js r161 and addons from jsDelivr; it is not offline self-contained.
  • The post-run weekly meter proves a shared Voxel+JRPG batch of 275 nominal Credits (range 250–300), but it cannot be split into task-level Credits.
  • No formal RemakeBench quality score or blind-voting result has been recorded.
Visible evidence gaps
  • 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.

RemakeBenchResearch console