Lava lamp v1 · Claude Code · run b79ff6d9 · 09.11.2026 21:58 Tallinn

Claude Fable 5.2

Medium effort

Works

8 of 8 checks passed. All 5 blocker checks passed.

What it built: 3 captures

Desktop capture of Claude Code (Fable 5.2)'s output on Lava lamp v1, medium effort, 09.11.2026 21:58 (run b79ff6d9)
Desktop · 501.1 KB · 705b32438b0d…
Phone capture of Claude Code (Fable 5.2)'s output on Lava lamp v1, medium effort, 09.11.2026 21:58 (run b79ff6d9)
Phone · 151.2 KB · bf8633ad4542…
Motion capture of Claude Code (Fable 5.2)'s output on Lava lamp v1, medium effort, 09.11.2026 21:58 (run b79ff6d9)
Motion · 500.8 KB · f1ef2e387354…

Along the row

Claude Fable 5.2 at every effort.

  1. Claude Fable 5.2 at low effort: desktop capture of run 96e988f7 Low4m 43s

    Works

    1 run
  2. Claude Fable 5.2 at medium effort: desktop capture of run b79ff6d9 Medium7m 47s

    Works

    This effort
  3. Claude Fable 5.2 at high effort: desktop capture of run 45666ded High34m 55s

    Works

    1 run
  4. Claude Fable 5.2 at extra high effort: desktop capture of run e47eb2fe Extra high40m 03s

    Works

    1 run
  5. MaxNot run

Down the column

Other models at medium effort.

Fastest median first · this run: 7m 47s

  1. Muse Ember at medium effort: desktop capture of run 44255a11Muse Ember32.2 s

    Works

    14.5× faster than this run
  2. GPT-6.1 Terra at medium effort: desktop capture of run f2424202GPT-6.1 Terra47.9 s

    Works

    9.8× faster than this run
  3. GPT-5.6 Luna at medium effort: desktop capture of run 69d0eb8eGPT-5.6 Luna59.1 s

    Works

    7.9× faster than this run
  4. GPT-6 Luna at medium effort: desktop capture of run 0e2e690dGPT-6 Luna1m 21s

    Works

    5.8× faster than this run
  5. GPT-5.6 Terra at medium effort: desktop capture of run 69024137GPT-5.6 Terra1m 48s

    Works

    4.3× faster than this run
  6. GPT-5.6 Sol at medium effort: desktop capture of run d3591876GPT-5.6 Sol2m 20s

    Fails

    3.3× faster than this run
  7. GPT-6 Astra at medium effort: desktop capture of run d0b2c47fGPT-6 Astra4m 15s

    Works

    1.8× faster than this run
  8. Claude Sonnet 5.5 at medium effort: desktop capture of run 2152d2b4Claude Sonnet 5.54m 44s

    Works

    1.6× faster than this run
  9. GPT-6.1 Sol at medium effort: desktop capture of run 0d942441GPT-6.1 Sol4m 52s

    Works

    1.6× faster than this run
  10. Claude Fable 5.1 at medium effort: desktop capture of run 6d90c6ceClaude Fable 5.14m 53s

    Works

    1.6× faster than this run
  11. Claude Haiku 5.5 at medium effort: desktop capture of run 012cc119Claude Haiku 5.55m 06s

    Works

    1.5× faster than this run
  12. GPT-6.2 Sol at medium effort: desktop capture of run 992d5562GPT-6.2 Sol7m 02s

    Works

    1.1× faster than this run
  13. Claude Opus 5.5 at medium effort: desktop capture of run 483f29b0Claude Opus 5.58m 56s

    Works

    1.1× slower than this run
  14. GPT-6.2 Astra at medium effort: desktop capture of run 136194bcGPT-6.2 Astra9m 42s

    Works

    1.2× slower than this run
  15. GPT-6 Sol at medium effort: desktop capture of run 18de4544GPT-6 Sol—

    Works

Hidden checks

8 of 8 checks passed.

  • Self-contained index.html is presentblocker
  • Visible page contentblocker
  • Visible autonomous animationblocker
  • Accessible on/off controlextra
  • 390 px mobile layout has no horizontal overflowextra
  • Local animation frame performanceextra
  • No browser errorsblocker
  • No attempted external or linked-resource requestsblocker

It works when every blocker passes. How it looks is reviewed blind: pending.

Run detailsagent config, what ran, tokens, source and revisions
  • Video (video): not published, videos are withheld. SHA-256 d20768c8c7fab359c38195c52e087cbaf91fafe134421fc072e95e326188b44d, 1164.8 KB.
Run ID
b79ff6d9-ae69-48f6-af0d-b9909af21541
Benchmark
lava-lamp v1
Experiment
Single run
Started
09.11.2026 21:58 Tallinn
Graded
revision 1 · 09.11.2026 22:07

Works and checks, in full

All 5 blocker checks passed.

The rule: a run works when every blocker check passes. If any blocker fails it does not work. If none fails but one has no result yet, it stays pending, never counted as a pass or a fail. Extra checks are reported but do not decide it. Works is not a judgement of looks: that is the visual review.

This is a legacy contract v1 run: the task has no public requirements, and these are the original lava-lamp checks, not a versioned hidden verifier. It is never compared with contract v2 runs.

Contract v1 checks for this run, with blocker and result
RequirementCheckBlockerResult
none (v1)Self-contained index.html is presentBlockerpass
none (v1)Visible page contentBlockerpass
none (v1)Visible autonomous animationBlockerpass
none (v1)Accessible on/off controlExtrapass
none (v1)390 px mobile layout has no horizontal overflowExtrapass
none (v1)Local animation frame performanceExtrapass
none (v1)No browser errorsBlockerpass
none (v1)No attempted external or linked-resource requestsBlockerpass

Verification completed; eligible for benchmark results.

Visual review

Pending.

Pending: the captures have not been reviewed blind yet. Checks decide whether it works; how it looks is decided only in that review.

What ran

Requested and detected.

Detected values come from the run’s own output, never from our settings.

Requested agent config beside what the run reported
FieldRequestedDetectedEvidence
Modelclaude-fable-5-2claude-fable-5-2harness result, 8 of 8 API requests
Effortmediummedium8 of 8 API requests
Harnessnot recorded2.1.321 (Claude Code)Version as the CLI reported it; binary SHA-256 abc9feeeca80…
Deadline60 min8m 46s usedCompleted

Tool decisions

15 calls · 14 accepted · 1 rejected

Each tool call is allowed or blocked by the harness’s permission settings before it runs. Allowed is not the same as succeeded: an allowed call can still fail.

  • Bash7 calls · 6 accepted · 1 rejected · 6 failed
  • Edit6 calls · 6 accepted
  • Write2 calls · 2 accepted

Telemetry

Time and usage.

Generation time
7m 47s
First content
1.0 s
Output speed
94.6 tokens/s, whole run including tools
API-equivalent cost
$2.71 (computed from the run’s token counts at list prices, not a charge)
Actual charge
not reported

Input tokens

397,412 tokens
  • Cache reads342,94686%
  • Cache writes54,45014%
  • Uncached input16<0.1%

Output tokens

44,211 tokens
  • Reasoning25,62358%
  • Other output18,58842%

Usage as the agent reported it at the end of the run. Cache reads are part of input; reasoning is part of output.

Transcript and source

Plain-text downloads.

The redacted transcript and the frozen output bundle, with LF line endings. Hashes are of the original files.

Redacted agent transcript: 56.0 KB, 13 redactions. Download transcript.txt

Source files

1 file · 25.8 KB
Files in the output bundle
FileSizeSHA-256Redactions
index.html25.8 KBe81e1a47d1c06ed54be42556437b30bdb686e68746542f8c605c284bfcc9d0ce0
Bundle SHA-256 e81e1a47d1c06ed54be42556437b30bdb686e68746542f8c605c284bfcc9d0ce

Revision history

This page shows revision 1.

Grades are append-only: a regrade adds a revision and never rewrites an earlier one.

Grading revisions of this run
RevisionGradedWorks
1 (current)09.11.2026 22:07works

Redaction: automated scan run; checked by Jaanus on 09.11.2026 22:10. Redacted: 4 local-path, 3 user-path, 6 username.