01HTML
MODEL ARCHIVE · M02 · Meta Superintelligence Labs
● 20 archived worksMuse Spark 1.3
This run used the contributor-free tier (providerID=opencode, reasoning variant xhigh) through opencode 1.18.29; harness and model identity are recorded separately from the model entity.
SELECTED WORKS
View all tasks →Selected works
Works first, numbers second.
ARCHIVE STATISTICS
Archive statistics
Runs20
Total duration1:10:47
API calls181
Tool calls416
Failures / retries12
Total tokens4.4M
Batch
Source report ↗Muse Spark 1.3 · Phase 1
Tasks15
Sessions15
Total duration25:45
API calls181
Tool calls178
Failures / retries8
Input tokens384,338
Output tokens154,563
Cache read3,130,484
Cost¥0
Batch
Source report ↗Muse Spark 1.3 · Phase 2 (Task 16–20, partial)
本批次为部分汇总：由已归档的 5 个 run 的逐题指标求和，不代表整个 phase-02（Task 16–45）；任一题缺失的指标记 null，不按 0 补齐。
Tasks5
Sessions5
Total duration45:02
API callsNot recorded
Tool calls238
Failures / retries4
Input tokens451,558
Output tokens146,779
Cache read20,144,127
CostNot recorded
PHASE PROGRESS
Phase progress
0102
PHASE-01
Phase 1 · 15 benchmark tasks
SVG illustration, single-file interactive apps, and multi-file web applications.
15 / 15 Tasks · Completed
15 RunsPHASE-02
Phase 2 · advanced creation tasks (Tasks 16–45)
In-browser animation, simulation, engineering and authoring tools. This archive currently covers only the tested and archived Tasks 16–20 (5 tasks × 2 models = 10 runs); the remaining tasks have not been run.
5 / 30 Tasks · In progress
5 RunsSNAPSHOTS
Model snapshots (harness-reported model id)
The model snapshot this run actually used; different snapshots of one model are not the same configuration.
Currentcontributor-free tier (xhigh)
muse-spark-1.3-contributor-freeHarness-reported value, recorded in each Phase 2 submission.json (opencode 1.18.30, desktop); the phase-01 flat-layout archive does not record a model id.CRITICAL NOTES
Phase assessment
AI phase assessments, simple average 86.5/100. The rubrics differ, so every score, method and disclaimer is kept side by side below; the average is an overview only and must not be turned into a ranking.
88.1/100
AI 评价，非盲评（会话工作目录即本仓库，可访问其他档案与人工评价）
Codex Desktop 0.153.4 / GPT-5 — non-blind, own rubric; not comparable to other reviewers or to the Phase 1 DeepSeek AI report.Source report ↗
84.9/100
AI 评价，非盲评（评价者此前已阅读第一阶段 DeepSeek 档案与题目文档）
维护 agent（DeepSeek Harness） — non-blind, own rubric; not comparable to other reviewers or to the Phase 1 DeepSeek AI report.Source report ↗
ALL WORKS
All works
01HTML
02SVG
03SVG
04SVG
05SVG
06SVG
07SVG
08HTML
09HTML
10HTML
11HTML
12HTML
13HTML
14HTML
15HTML
16SVG
17HTML
18HTML
19HTML
20HTML
RUN HISTORY
Run history
One row per archived run; open a row for the full record.
ProviderMeta Superintelligence Labs
Works20
AI average86.5/100（2）
Archive commit
b8d0235e89ac…