Launch 005 · Four Flash models

Qwen 3.8 Flash vs DeepSeek V4.1 Flash vs GLM 5.3 Flash vs Gemini 3.8 Flash

There was no universal winner. Flash-model performance was brutally task-specific.

Fifty-four direct model clips now carry the comparison. Qwen led several game and 3D-scene rounds, DeepSeek produced the better submitted CAD result, GLM alone passed Apple Stem, and Gemini led several robot tasks.

Editorial preferences are not aggregate scores. Single attempts do not establish reliability.

Watch every fixture Inspect all public results

14direct comparison fixtures
54model clips
54permanent results
35existing records reused
19campaign records added
No scoreno universal ranking

14 fixtures · 54 direct clips · poster first

The outputs are the main event.

Only the visible fixture mounts video. Offscreen comparisons remain posters, playback pauses outside the viewport, and reduced-motion visitors keep the poster instead of autoplaying footage.

Fixture 01 · episode 00:18

Rubik’s Cube single turn

Dex Cube · Rubik’s Cube · Rubiks Cube · Cube Turn

Editorial observationGLM and Gemini completed the single-turn task; Gemini was fastest. Qwen and DeepSeek did not complete it.Pass/fail labels are video-campaign judgments. No machine-readable robot validator was supplied.

Qwen 3.8 Flash · dex-cube-turn presentation recording poster — Campaign evidence
Qwen 3.8 Flash
Open result
Displayed cost
$0.2616
Wall clock
48m 34.9s
Status
campaign evidence not harness ledger
Reasoning
Max
**Validation scope.** Campaign judgment: did not pass
**Material caveat.** Campaign-evidence record; this attempt was not generated from the public Harness ledger.
Clip metric vs result-receipt scope
This permanent campaign-evidence record uses the handoff’s exact displayed cost and wall clock; it is not a generated Harness ledger.
Public record identity: Qwen 3.8 Flash · Qwen 3.8 Flash · Unavailable · Unavailable

DeepSeek V4.1 Flash · dex-cube-turn presentation recording poster — Campaign evidence
DeepSeek V4.1 Flash
Open result
Displayed cost
$0.0957
Wall clock
16m 38.0s
Status
campaign evidence not harness ledger
Reasoning
Max
**Validation scope.** Campaign judgment: did not pass
**Material caveat.** Campaign-evidence record; this attempt was not generated from the public Harness ledger.
Clip metric vs result-receipt scope
This permanent campaign-evidence record uses the handoff’s exact displayed cost and wall clock; it is not a generated Harness ledger.
Public record identity: DeepSeek V4.1 Flash · DeepSeek V4.1 Flash · Unavailable · Unavailable

GLM 5.3 Flash · dex-cube-turn presentation recording poster — Campaign evidence
GLM 5.3 Flash
Open result
Displayed cost
$0.3414
Wall clock
40m 47.8s
Status
campaign evidence not harness ledger
Reasoning
Max
**Validation scope.** Campaign judgment: pass
**Material caveat.** Campaign-evidence record; this attempt was not generated from the public Harness ledger.
Clip metric vs result-receipt scope
This permanent campaign-evidence record uses the handoff’s exact displayed cost and wall clock; it is not a generated Harness ledger.
Public record identity: GLM 5.3 Flash · GLM 5.3 Flash · Unavailable · Unavailable

Gemini 3.8 Flash · dex-cube-turn presentation recording poster — Campaign evidence
Gemini 3.8 Flash
Open result
Displayed cost
$2.09
Wall clock
15m 35s
Status
campaign evidence not harness ledger
Reasoning
High
**Validation scope.** Campaign judgment: pass
**Material caveat.** Campaign-evidence record; this attempt was not generated from the public Harness ledger.
Clip metric vs result-receipt scope
This permanent campaign-evidence record uses the handoff’s exact displayed cost and wall clock; it is not a generated Harness ledger.
Public record identity: Gemini 3.8 Flash · Gemini 3.8 Flash · Unavailable · Unavailable

Fixture 02 · episode 00:18

Rubik’s Cube three-move extension

Dex Cube · Rubik’s Cube · Rubiks Cube · Three Move Cube

Editorial observationNeither GLM nor Gemini completed the harder three-move extension.Only the two single-turn passers advanced. Qwen and DeepSeek have no three-move attempt and are shown as did not advance, not as missing or failed runs.

GLM 5.3 Flash · dex-cube-three-move presentation recording poster — Campaign evidence
GLM 5.3 Flash
Open result
Displayed cost
$1.4412
Wall clock
1h 31m 15.1s
Status
campaign evidence not harness ledger
Reasoning
Max
**Validation scope.** Campaign judgment: did not complete
**Material caveat.** Campaign-evidence record; this attempt was not generated from the public Harness ledger.
Clip metric vs result-receipt scope
This permanent campaign-evidence record uses the handoff’s exact displayed cost and wall clock; it is not a generated Harness ledger.
Public record identity: GLM 5.3 Flash · GLM 5.3 Flash · Unavailable · Unavailable

Gemini 3.8 Flash · dex-cube-three-move presentation recording poster — Campaign evidence
Gemini 3.8 Flash
Open result
Displayed cost
$5.65
Wall clock
44m 30s
Status
campaign evidence not harness ledger
Reasoning
High
**Validation scope.** Campaign judgment: did not complete
**Material caveat.** Campaign-evidence record; this attempt was not generated from the public Harness ledger.
Clip metric vs result-receipt scope
This permanent campaign-evidence record uses the handoff’s exact displayed cost and wall clock; it is not a generated Harness ledger.
Public record identity: Gemini 3.8 Flash · Gemini 3.8 Flash · Unavailable · Unavailable

Qwen 3.8 FlashDid not advance

Did not advance after the single-turn task; no three-move attempt exists.

DeepSeek V4.1 FlashDid not advance

Did not advance after the single-turn task; no three-move attempt exists.

Fixture 03 · episode 01:31

Campfire

Campfire · Campfire Under a Starry Night

Editorial observationQwen produced one of the strongest visual outputs in the campaign’s Campfire fixture.Visual preference is editorial. Each linked ledger retains its own validation state and metric scope.

Qwen 3.8 Flash · campfire presentation recording poster — Existing public result
Qwen 3.8 Flash
Open result
Displayed cost
$0.1877
Wall clock
55m 21.924s
Status
PENDING JUDGE
Reasoning
Unreported
**Validation scope.** PENDING_JUDGE
**Material caveat.** No additional public caveat was supplied by this result record.
Clip metric vs result-receipt scope
Clip handoff: $0.1877, 55m 21.924s. Existing result receipt: $0.19 Provider-recorded OpenRouter cost; Not recorded. More precise clip formatting is preserved without changing the receipt.
Public record identity: Qwen3.8 Flash · qwen3.8-flash-openrouter · Provider not separately recorded

DeepSeek V4.1 Flash · campfire presentation recording poster — Existing public result
DeepSeek V4.1 Flash
Open result
Displayed cost
$0.5276
Wall clock
1h 10m 27.4s
Status
PASS
Reasoning
Vision exp
**Validation scope.** PASS
**Material caveat.** No additional public caveat was supplied by this result record.
Clip metric vs result-receipt scope
Metric scopes materially disagree. Clip handoff: $0.5276, 1h 10m 27.4s. Existing result receipt: $0.08 Provider-recorded OpenRouter cost; Not recorded. Both remain visible; neither silently replaces the other.
Public record identity: DeepSeek V4 Flash · deepseek-v4-flash-vision-exp-openrouter · Provider not separately recorded

GLM 5.3 Flash · campfire presentation recording poster — Existing public result
GLM 5.3 Flash
Open result
Displayed cost
$0.0151
Wall clock
7m 21.718s
Status
PASS
Reasoning
Max
**Validation scope.** PASS
**Material caveat.** No additional public caveat was supplied by this result record.
Clip metric vs result-receipt scope
Clip handoff: $0.0151, 7m 21.718s. Existing result receipt: $0.02 Provider-recorded OpenRouter cost; Not recorded. More precise clip formatting is preserved without changing the receipt.
Public record identity: GLM 5.3 Flash · glm-5.3-flash-max-opencode · Provider not separately recorded

Gemini 3.8 Flash · campfire presentation recording poster — Existing public result
Gemini 3.8 Flash
Open result
Displayed cost
$4.44
Wall clock
1h 11m 27.5s
Status
complete token timing artifact and validation ledger
Reasoning
Highest
**Validation scope.** pass
**Material caveat.** No blind visual quality score is attached.
Clip metric vs result-receipt scope
Clip handoff: $4.44, 1h 11m 27.5s. Existing result receipt: $4.44 Provider-recorded usage estimate, not an itemized subscription cash charge; 4287.5s wall-clock. More precise clip formatting is preserved without changing the receipt.
Public record identity: Gemini 3.8 Flash · Gemini 3.8 Flash · OpenRouter · OpenCode 1.18.23

Fixture 04 · episode 01:31

Courtyard Blossom Stage 2 / Blender arena

Fighter Blossom · Courtyard Blossom · Blender Arena

Editorial observationQwen produced one of the strongest arena outputs in the campaign.The handoff calls this Fighter Blossom; the permanent benchmark remains Courtyard Blossom Stage 2.

Qwen 3.8 Flash · courtyard-blossom-stage-2 presentation recording poster — Existing public result
Qwen 3.8 Flash
Open result
Displayed cost
$1.17
Wall clock
2h 35m 9.1s
Status
FAIL
Reasoning
Unreported
**Validation scope.** FAIL
**Material caveat.** No additional public caveat was supplied by this result record.
Clip metric vs result-receipt scope
Metric scopes materially disagree. Clip handoff: $1.17, 2h 35m 9.1s. Existing result receipt: $0.71 Provider-recorded OpenRouter cost; Not recorded. Both remain visible; neither silently replaces the other.
Public record identity: Qwen3.8 Flash · qwen3.8-flash-openrouter · Provider not separately recorded

DeepSeek V4.1 Flash · courtyard-blossom-stage-2 presentation recording poster — Existing public result
DeepSeek V4.1 Flash
Open result
Displayed cost
$0.3613
Wall clock
59m 59.2s
Status
PARTIAL
Reasoning
Vision exp
**Validation scope.** PARTIAL
**Material caveat.** No additional public caveat was supplied by this result record.
Clip metric vs result-receipt scope
Metric scopes materially disagree. Clip handoff: $0.3613, 59m 59.2s. Existing result receipt: $0.46 Provider-recorded OpenRouter cost; Not recorded. Both remain visible; neither silently replaces the other.
Public record identity: DeepSeek V4 Flash · deepseek-v4-flash-vision-exp-openrouter · Provider not separately recorded

GLM 5.3 Flash · courtyard-blossom-stage-2 presentation recording poster — Existing public result
GLM 5.3 Flash
Open result
Displayed cost
$0.7893
Wall clock
1h 37m 48.2s
Status
FAIL
Reasoning
Max
**Validation scope.** FAIL
**Material caveat.** No additional public caveat was supplied by this result record.
Clip metric vs result-receipt scope
Metric scopes materially disagree. Clip handoff: $0.7893, 1h 37m 48.2s. Existing result receipt: $0.68 Provider-recorded OpenRouter cost; Not recorded. Both remain visible; neither silently replaces the other.
Public record identity: GLM 5.3 Flash · glm-5.3-flash-max-opencode · Provider not separately recorded

Gemini 3.8 Flash · courtyard-blossom-stage-2 presentation recording poster — Existing public result
Gemini 3.8 Flash
Open result
Displayed cost
$5.02
Wall clock
47m 18.7s
Status
complete token timing artifact and validation ledger
Reasoning
Highest
**Validation scope.** pass_with_visual_caveat
**Material caveat.** No blind visual quality score is attached.
Clip metric vs result-receipt scope
Clip handoff: $5.02, 47m 18.7s. Existing result receipt: $5.02 Provider-recorded usage estimate, not an itemized subscription cash charge; 2838.7s wall-clock. More precise clip formatting is preserved without changing the receipt.
Public record identity: Gemini 3.8 Flash · Gemini 3.8 Flash · OpenRouter · OpenCode 1.18.23

Fixture 05 · episode 02:35

Interactive voxel diorama

Interactive Voxel Diorama · Voxel Diorama · Voxel Map Viewer

Editorial observationQwen was the editorial favorite in the interactive voxel-diorama round.The clips are direct presentation evidence, not standardized FPS or hardware evidence.

Qwen 3.8 Flash · interactive-voxel-diorama presentation recording poster — Existing public result
Qwen 3.8 Flash
Open result
Displayed cost
$0.3338
Wall clock
1h 12m 56.365s
Status
PASS
Reasoning
Unreported
**Validation scope.** PASS
**Material caveat.** No additional public caveat was supplied by this result record.
Clip metric vs result-receipt scope
Clip handoff: $0.3338, 1h 12m 56.365s. Existing result receipt: $0.33 Provider-recorded OpenRouter cost; Not recorded. More precise clip formatting is preserved without changing the receipt.
Public record identity: Qwen3.8 Flash · qwen3.8-flash-openrouter · Provider not separately recorded

DeepSeek V4.1 Flash · interactive-voxel-diorama presentation recording poster — Existing public result
DeepSeek V4.1 Flash
Open result
Displayed cost
$0.4925
Wall clock
1h 12m 32.3s
Status
PASS
Reasoning
Vision exp
**Validation scope.** PASS
**Material caveat.** No additional public caveat was supplied by this result record.
Clip metric vs result-receipt scope
Metric scopes materially disagree. Clip handoff: $0.4925, 1h 12m 32.3s. Existing result receipt: $0.18 Provider-recorded OpenRouter cost; Not recorded. Both remain visible; neither silently replaces the other.
Public record identity: DeepSeek V4 Flash · deepseek-v4-flash-vision-exp-openrouter · Provider not separately recorded

GLM 5.3 Flash · interactive-voxel-diorama presentation recording poster — Existing public result
GLM 5.3 Flash
Open result
Displayed cost
$0.5224
Wall clock
1h 52m 57.216s
Status
PASS
Reasoning
Max
**Validation scope.** PASS
**Material caveat.** No additional public caveat was supplied by this result record.
Clip metric vs result-receipt scope
Clip handoff: $0.5224, 1h 52m 57.216s. Existing result receipt: $0.52 Provider-recorded OpenRouter cost; Not recorded. More precise clip formatting is preserved without changing the receipt.
Public record identity: GLM 5.3 Flash · glm-5.3-flash-max-opencode · Provider not separately recorded

Gemini 3.8 Flash · interactive-voxel-diorama presentation recording poster — Existing public result
Gemini 3.8 Flash
Open result
Displayed cost
$0.5327
Wall clock
2h 2m 44.5s
Status
complete token timing artifact and validation ledger
Reasoning
Highest
**Validation scope.** pass
**Material caveat.** No blind visual quality score is attached.
Clip metric vs result-receipt scope
Clip handoff: $0.5327, 2h 2m 44.5s. Existing result receipt: $0.53 Provider-recorded usage estimate, not an itemized subscription cash charge; 7364.5s wall-clock. More precise clip formatting is preserved without changing the receipt.
Public record identity: Gemini 3.8 Flash · Gemini 3.8 Flash · OpenRouter · OpenCode 1.18.23

Fixture 06 · episode 03:06

JRPG boss battle

JRPG · JRPG Boss Battle

Editorial observationThe JRPG attempts split across completion, validation, and visual quality rather than producing one universal model ranking.Each exact public ledger remains authoritative for its validator and missing-evidence claims.

Qwen 3.8 Flash · jrpg-boss-battle presentation recording poster — Existing public result
Qwen 3.8 Flash
Open result
Displayed cost
$0.1398
Wall clock
36m 43.822s
Status
PENDING JUDGE
Reasoning
Unreported
**Validation scope.** PENDING_JUDGE
**Material caveat.** No additional public caveat was supplied by this result record.
Clip metric vs result-receipt scope
Clip handoff: $0.1398, 36m 43.822s. Existing result receipt: $0.14 Provider-recorded OpenRouter cost; Not recorded. More precise clip formatting is preserved without changing the receipt.
Public record identity: Qwen3.8 Flash · qwen3.8-flash-openrouter · Provider not separately recorded

DeepSeek V4.1 Flash · jrpg-boss-battle presentation recording poster — Existing public result
DeepSeek V4.1 Flash
Open result
Displayed cost
$0.1561
Wall clock
34m 6.7s
Status
PASS
Reasoning
Vision exp
**Validation scope.** PASS
**Material caveat.** No additional public caveat was supplied by this result record.
Clip metric vs result-receipt scope
Metric scopes materially disagree. Clip handoff: $0.1561, 34m 6.7s. Existing result receipt: $0.23 Provider-recorded OpenRouter cost; Not recorded. Both remain visible; neither silently replaces the other.
Public record identity: DeepSeek V4 Flash · deepseek-v4-flash-vision-exp-openrouter · Provider not separately recorded

GLM 5.3 Flash · jrpg-boss-battle presentation recording poster — Existing public result
GLM 5.3 Flash
Open result
Displayed cost
$0.1571
Wall clock
52m 42.980s
Status
PARTIAL
Reasoning
Max
**Validation scope.** PARTIAL
**Material caveat.** No additional public caveat was supplied by this result record.
Clip metric vs result-receipt scope
Clip handoff: $0.1571, 52m 42.980s. Existing result receipt: $0.16 Provider-recorded OpenRouter cost; Not recorded. More precise clip formatting is preserved without changing the receipt.
Public record identity: GLM 5.3 Flash · glm-5.3-flash-max-opencode · Provider not separately recorded

Gemini 3.8 Flash · jrpg-boss-battle presentation recording poster — Existing public result
Gemini 3.8 Flash
Open result
Displayed cost
$1.12
Wall clock
25m 30.7s
Status
complete token timing artifact and validation ledger
Reasoning
Highest
**Validation scope.** pass
**Material caveat.** No blind visual quality score is attached.
Clip metric vs result-receipt scope
Clip handoff: $1.12, 25m 30.7s. Existing result receipt: $1.12 Provider-recorded usage estimate, not an itemized subscription cash charge; 1530.7s wall-clock. More precise clip formatting is preserved without changing the receipt.
Public record identity: Gemini 3.8 Flash · Gemini 3.8 Flash · OpenRouter · OpenCode 1.18.23

Fixture 07 · episode 03:06

Neon Gunner

Editorial observationNeon Gunner again separated visual preference from recorded functional and validation states.Editorial footage does not replace the permanent result ledger.

DeepSeek V4.1 Flash · neon-gunner presentation recording poster — Existing public result
DeepSeek V4.1 Flash
Open result
Displayed cost
$0.3436
Wall clock
53m 44.4s
Status
PARTIAL
Reasoning
Vision exp
**Validation scope.** PARTIAL
**Material caveat.** No additional public caveat was supplied by this result record.
Clip metric vs result-receipt scope
Metric scopes materially disagree. Clip handoff: $0.3436, 53m 44.4s. Existing result receipt: $0.54 Provider-recorded OpenRouter cost; Not recorded. Both remain visible; neither silently replaces the other.
Public record identity: DeepSeek V4 Flash · deepseek-v4-flash-vision-exp-openrouter · Provider not separately recorded

GLM 5.3 Flash · neon-gunner presentation recording poster — Existing public result
GLM 5.3 Flash
Open result
Displayed cost
$0.7926
Wall clock
3h 9m 20.539s
Status
PENDING JUDGE
Reasoning
Max
**Validation scope.** PENDING_JUDGE
**Material caveat.** No additional public caveat was supplied by this result record.
Clip metric vs result-receipt scope
Clip handoff: $0.7926, 3h 9m 20.539s. Existing result receipt: $0.79 Provider-recorded OpenRouter cost; Not recorded. More precise clip formatting is preserved without changing the receipt.
Public record identity: GLM 5.3 Flash · glm-5.3-flash-max-opencode · Provider not separately recorded

Gemini 3.8 Flash · neon-gunner presentation recording poster — Existing public result
Gemini 3.8 Flash
Open result
Displayed cost
$2.59
Wall clock
24m 23.9s
Status
complete token timing artifact and validation ledger
Reasoning
Highest
**Validation scope.** pass_with_nonbenchmark_caveat
**Material caveat.** No blind visual quality score is attached.
Clip metric vs result-receipt scope
Clip handoff: $2.59, 24m 23.9s. Existing result receipt: $2.59 Provider-recorded usage estimate, not an itemized subscription cash charge; 1463.9s wall-clock. More precise clip formatting is preserved without changing the receipt.
Public record identity: Gemini 3.8 Flash · Gemini 3.8 Flash · OpenRouter · OpenCode 1.18.23

Fixture 08 · episode 04:03

Sekiro main menu

Sekiro · Sekiro Main Menu

Editorial observationThe menu recreations differ in motion, composition, and finish; their public records preserve the evidence gaps.Direct footage is presentation evidence, not a fabricated numerical score.

Qwen 3.8 Flash · sekiro-main-menu presentation recording poster — Existing public result
Qwen 3.8 Flash
Open result
Displayed cost
$0.4615
Wall clock
1h 50m 39.621s
Status
PARTIAL
Reasoning
Unreported
**Validation scope.** PARTIAL
**Material caveat.** No additional public caveat was supplied by this result record.
Clip metric vs result-receipt scope
Clip handoff: $0.4615, 1h 50m 39.621s. Existing result receipt: $0.46 Provider-recorded OpenRouter cost; Not recorded. More precise clip formatting is preserved without changing the receipt.
Public record identity: Qwen3.8 Flash · qwen3.8-flash-openrouter · Provider not separately recorded

DeepSeek V4.1 Flash · sekiro-main-menu presentation recording poster — Existing public result
DeepSeek V4.1 Flash
Open result
Displayed cost
$0.3317
Wall clock
58m 15.4s
Status
PARTIAL
Reasoning
Vision exp
**Validation scope.** PARTIAL
**Material caveat.** No additional public caveat was supplied by this result record.
Clip metric vs result-receipt scope
Metric scopes materially disagree. Clip handoff: $0.3317, 58m 15.4s. Existing result receipt: $0.49 Provider-recorded OpenRouter cost; Not recorded. Both remain visible; neither silently replaces the other.
Public record identity: DeepSeek V4 Flash · deepseek-v4-flash-vision-exp-openrouter · Provider not separately recorded

GLM 5.3 Flash · sekiro-main-menu presentation recording poster — Existing public result
GLM 5.3 Flash
Open result
Displayed cost
$0.106
Wall clock
35m 18.685s
Status
PARTIAL
Reasoning
Max
**Validation scope.** PARTIAL
**Material caveat.** No additional public caveat was supplied by this result record.
Clip metric vs result-receipt scope
Clip handoff: $0.106, 35m 18.685s. Existing result receipt: $0.11 Provider-recorded OpenRouter cost; Not recorded. More precise clip formatting is preserved without changing the receipt.
Public record identity: GLM 5.3 Flash · glm-5.3-flash-max-opencode · Provider not separately recorded

Gemini 3.8 Flash · sekiro-main-menu presentation recording poster — Existing public result
Gemini 3.8 Flash
Open result
Displayed cost
$0.7402
Wall clock
11m 56.6s
Status
complete token timing artifact and validation ledger
Reasoning
Highest
**Validation scope.** pass
**Material caveat.** No blind visual quality score is attached.
Clip metric vs result-receipt scope
Clip handoff: $0.7402, 11m 56.6s. Existing result receipt: $0.74 Provider-recorded usage estimate, not an itemized subscription cash charge; 716.6s wall-clock. More precise clip formatting is preserved without changing the receipt.
Public record identity: Gemini 3.8 Flash · Gemini 3.8 Flash · OpenRouter · OpenCode 1.18.23

Fixture 09 · episode 04:03

Space Flight Stage 1

Space Flight · Space Flight Stage 1

Editorial observationQwen’s Space Flight output was one of its strongest campaign showcases.Qwen now has an explicit campaign-evidence result. Existing DeepSeek, GLM, and Gemini ledgers keep their stronger committed provenance and conflicting metric scopes.

Qwen 3.8 Flash · space-flight-stage-1 presentation recording poster — Campaign evidence
Qwen 3.8 Flash
Open result
Displayed cost
$0.8747
Wall clock
2h 33m 50.9s
Status
campaign evidence not harness ledger
Reasoning
Unavailable
**Validation scope.** No machine-readable validation outcome supplied
**Material caveat.** Campaign-evidence record; this attempt was not generated from the public Harness ledger.
Clip metric vs result-receipt scope
This permanent campaign-evidence record uses the handoff’s exact displayed cost and wall clock; it is not a generated Harness ledger.
Public record identity: Qwen 3.8 Flash · Qwen 3.8 Flash · Unavailable · Unavailable

DeepSeek V4.1 Flash · space-flight-stage-1 presentation recording poster — Existing public result
DeepSeek V4.1 Flash
Open result
Displayed cost
$0.242
Wall clock
45m 44.4s
Status
partial token timing and artifact ledger
Reasoning
Max
**Validation scope.** not_run
**Material caveat.** This is a two-user-turn protocol deviation: the original benchmark prompt needed one ‘continue’ rescue/follow-up turn.
Clip metric vs result-receipt scope
Metric scopes materially disagree. Clip handoff: $0.242, 45m 44.4s. Existing result receipt: $0.25 API-equivalent usage accounting, not an itemized subscription cash charge; 5h 0m 0.3s wall-clock. Both remain visible; neither silently replaces the other.
Public record identity: DeepSeek V4 Flash · DeepSeek V4 Flash · OpenCode Go

GLM 5.3 Flash · space-flight-stage-1 presentation recording poster — Existing public result
GLM 5.3 Flash
Open result
Displayed cost
$1.07
Wall clock
3h 14m 24.2s
Status
partial multi turn token timing and artifact ledger
Reasoning
Max
**Validation scope.** No machine-readable validator attached
**Material caveat.** This is a two-turn, 13-user-message ZCode session, not a one-shot benchmark run.
Clip metric vs result-receipt scope
Metric scopes materially disagree. Clip handoff: $1.07, 3h 14m 24.2s. Existing result receipt: $4.46 UNVERIFIED reference scenario: GLM-5.2/GLM-5.1 official rates applied to GLM-5.3 token counts; not a verified GLM-5.3 charge or subscription marginal-cash invoice; 48m 14.1s wall-clock. Both remain visible; neither silently replaces the other.
Public record identity: GLM 5.3 Flash · glm-5.3-flash · ZCode

Gemini 3.8 Flash · space-flight-stage-1 presentation recording poster — Existing public result
Gemini 3.8 Flash
Open result
Displayed cost
$4.3
Wall clock
2h 26m 30.3s
Status
complete token timing artifact and validation ledger
Reasoning
Highest
**Validation scope.** pass_with_visual_caveat
**Material caveat.** No blind visual quality score is attached.
Clip metric vs result-receipt scope
Clip handoff: $4.3, 2h 26m 30.3s. Existing result receipt: $4.30 Provider-recorded usage estimate, not an itemized subscription cash charge; 8790.3s wall-clock. More precise clip formatting is preserved without changing the receipt.
Public record identity: Gemini 3.8 Flash · Gemini 3.8 Flash · OpenRouter · OpenCode 1.18.23

Fixture 10 · episode 05:00

Turbofan CAD

Turbofan · Turbofan Engine · CAD-001

Editorial observationDeepSeek delivered the better submitted CAD result. Gemini produced valid solids but did not complete the required delivery; Qwen timed out and GLM truncated without native geometry.Gemini’s operator-created assembly and presentation clip are post-run work, never model-completed scored motion evidence.

Qwen 3.8 Flash · turbofan-engine presentation recording poster — Campaign evidence
Qwen 3.8 Flash
Open result
Displayed cost
$0.355312
Wall clock
1h timeout
Status
campaign evidence not harness ledger
Reasoning
Max
**Validation scope.** 0 parts · 0 assembly instances · 0 mates
**Material caveat.** Campaign-evidence record; this attempt was not generated from the public Harness ledger.
Clip metric vs result-receipt scope
This permanent campaign-evidence record uses the handoff’s exact displayed cost and wall clock; it is not a generated Harness ledger.
Public record identity: Qwen 3.8 Flash · openrouter/qwen/qwen3.8-flash · OpenRouter · OpenCode 1.18.23

DeepSeek V4.1 Flash · turbofan-engine presentation recording poster — Campaign evidence
DeepSeek V4.1 Flash
Open result
Displayed cost
$0.375637
Wall clock
55m 52.064s
Status
campaign evidence not harness ledger
Reasoning
Max
**Validation scope.** Three valid native solids; wrong-sign coupling and LP/HP interference
**Material caveat.** Campaign-evidence record; this attempt was not generated from the public Harness ledger.
Clip metric vs result-receipt scope
This permanent campaign-evidence record uses the handoff’s exact displayed cost and wall clock; it is not a generated Harness ledger.
Public record identity: DeepSeek V4.1 Flash · opencode-go/deepseek-v4.1-flash · OpenCode Go · OpenCode 1.18.23

GLM 5.3 Flash · turbofan-engine presentation recording poster — Campaign evidence
GLM 5.3 Flash
Open result
Displayed cost
$0.215242
Wall clock
50m 37s
Status
campaign evidence not harness ledger
Reasoning
Max
**Validation scope.** No source or native geometry delivered
**Material caveat.** Campaign-evidence record; this attempt was not generated from the public Harness ledger.
Clip metric vs result-receipt scope
This permanent campaign-evidence record uses the handoff’s exact displayed cost and wall clock; it is not a generated Harness ledger.
Public record identity: GLM 5.3 Flash · opencode-go/glm-5.3-flash · OpenCode Go · OpenCode 1.18.23

Gemini 3.8 Flash · turbofan-engine presentation recording poster — Campaign evidence
Gemini 3.8 Flash
Open result
Displayed cost
$3.793750
Wall clock
1h 0m 4.361s
Status
campaign evidence not harness ledger
Reasoning
High
**Validation scope.** Four valid native solids from compiling FeatureScript
**Material caveat.** Campaign-evidence record; this attempt was not generated from the public Harness ledger.
Clip metric vs result-receipt scope
This permanent campaign-evidence record uses the handoff’s exact displayed cost and wall clock; it is not a generated Harness ledger.
Public record identity: Gemini 3.8 Flash · gemini-3.8-flash-high · Google · Antigravity CLI

Fixture 11 · episode 05:35

Neo-Gothic Storm City

Neo-Gothic Storm City · Neo Gothic City

Editorial observationDeepSeek was among the strongest graphics models in the Neo-Gothic city fixture.Visual preference remains editorial; linked ledgers retain exact completion and validation states.

Qwen 3.8 Flash · neo-gothic-storm-city presentation recording poster — Existing public result
Qwen 3.8 Flash
Open result
Displayed cost
$0.2038
Wall clock
47m 43.138s
Status
PARTIAL
Reasoning
Unreported
**Validation scope.** PARTIAL
**Material caveat.** No additional public caveat was supplied by this result record.
Clip metric vs result-receipt scope
Clip handoff: $0.2038, 47m 43.138s. Existing result receipt: $0.20 Provider-recorded OpenRouter cost; Not recorded. More precise clip formatting is preserved without changing the receipt.
Public record identity: Qwen3.8 Flash · qwen3.8-flash-openrouter · Provider not separately recorded

DeepSeek V4.1 Flash · neo-gothic-storm-city presentation recording poster — Existing public result
DeepSeek V4.1 Flash
Open result
Displayed cost
$0.1596
Wall clock
1h 21m 50.825s
Status
INCOMPLETE EXHAUSTED
Reasoning
Vision exp
**Validation scope.** INCOMPLETE_EXHAUSTED
**Material caveat.** No additional public caveat was supplied by this result record.
Clip metric vs result-receipt scope
Metric scopes materially disagree. Clip handoff: $0.1596, 1h 21m 50.825s. Existing result receipt: $0.42 Provider-recorded OpenRouter cost; Not recorded. Both remain visible; neither silently replaces the other.
Public record identity: DeepSeek V4 Flash · deepseek-v4-flash-vision-exp-openrouter · Provider not separately recorded

GLM 5.3 Flash · neo-gothic-storm-city presentation recording poster — Existing public result
GLM 5.3 Flash
Open result
Displayed cost
$0.1182
Wall clock
35m 58.906s
Status
PARTIAL
Reasoning
Max
**Validation scope.** PARTIAL
**Material caveat.** No additional public caveat was supplied by this result record.
Clip metric vs result-receipt scope
Metric scopes materially disagree. Clip handoff: $0.1182, 35m 58.906s. Existing result receipt: $0.45 Provider-recorded OpenRouter cost; Not recorded. Both remain visible; neither silently replaces the other.
Public record identity: GLM 5.3 Flash · glm-5.3-flash-max-opencode · Provider not separately recorded

Gemini 3.8 Flash · neo-gothic-storm-city presentation recording poster — Existing public result
Gemini 3.8 Flash
Open result
Displayed cost
$2.59
Wall clock
47m 47.906s
Status
complete token timing artifact and validation ledger
Reasoning
Highest
**Validation scope.** pass
**Material caveat.** No blind visual quality score is attached.
Clip metric vs result-receipt scope
Clip handoff: $2.59, 47m 47.906s. Existing result receipt: $2.59 Provider-recorded usage estimate, not an itemized subscription cash charge; 2867.9s wall-clock. More precise clip formatting is preserved without changing the receipt.
Public record identity: Gemini 3.8 Flash · Gemini 3.8 Flash · OpenRouter · OpenCode 1.18.23

Fixture 12 · episode 05:35

Infinite Cathedral

Infinite Cathedral · Cathedral Shader

Editorial observationThe Infinite Cathedral attempts split by visual execution, recorded completion, and evidence quality.No clip-level preference becomes an aggregate score or universal winner.

Qwen 3.8 Flash · infinite-cathedral presentation recording poster — Existing public result
Qwen 3.8 Flash
Open result
Displayed cost
$0.2618
Wall clock
59m 32.948s
Status
PARTIAL
Reasoning
Unreported
**Validation scope.** PARTIAL
**Material caveat.** No additional public caveat was supplied by this result record.
Clip metric vs result-receipt scope
Clip handoff: $0.2618, 59m 32.948s. Existing result receipt: $0.26 Provider-recorded OpenRouter cost; Not recorded. More precise clip formatting is preserved without changing the receipt.
Public record identity: Qwen3.8 Flash · qwen3.8-flash-openrouter · Provider not separately recorded

DeepSeek V4.1 Flash · infinite-cathedral presentation recording poster — Existing public result
DeepSeek V4.1 Flash
Open result
Displayed cost
$0.2177
Wall clock
35m 47.810s
Status
PENDING JUDGE
Reasoning
Vision exp
**Validation scope.** PENDING_JUDGE
**Material caveat.** No additional public caveat was supplied by this result record.
Clip metric vs result-receipt scope
Metric scopes materially disagree. Clip handoff: $0.2177, 35m 47.810s. Existing result receipt: $0.24 Provider-recorded OpenRouter cost; Not recorded. Both remain visible; neither silently replaces the other.
Public record identity: DeepSeek V4 Flash · deepseek-v4-flash-vision-exp-openrouter · Provider not separately recorded

GLM 5.3 Flash · infinite-cathedral presentation recording poster — Existing public result
GLM 5.3 Flash
Open result
Displayed cost
$0.5895
Wall clock
1h 26m 0.031s
Status
complete
Reasoning
Max
**Validation scope.** complete
**Material caveat.** The run required repeated user-message continuations because the OpenCode endpoint frequently exited or stalled before task completion.
Clip metric vs result-receipt scope
Clip handoff: $0.5895, 1h 26m 0.031s. Existing result receipt: Not recorded; Not recorded. More precise clip formatting is preserved without changing the receipt.
Public record identity: GLM 5.3 Flash · glm-5.3-flash · Provider not separately recorded · OpenCode CLI 1.18.18

Gemini 3.8 Flash · infinite-cathedral presentation recording poster — Existing public result
Gemini 3.8 Flash
Open result
Displayed cost
$2.33
Wall clock
1h 6m 37.863s
Status
complete token timing artifact and validation ledger
Reasoning
Highest
**Validation scope.** pass
**Material caveat.** No blind visual quality score is attached.
Clip metric vs result-receipt scope
Clip handoff: $2.33, 1h 6m 37.863s. Existing result receipt: $2.33 Provider-recorded usage estimate, not an itemized subscription cash charge; 3997.9s wall-clock. More precise clip formatting is preserved without changing the receipt.
Public record identity: Gemini 3.8 Flash · Gemini 3.8 Flash · OpenRouter · OpenCode 1.18.23

Fixture 13 · episode 06:20

Eiffel Tower drawing

Eiffel Tower · Eiffel Tower Drawing · Robot Drawing

Editorial observationGemini was the editorial winner and the fastest shown Eiffel Tower attempt at 8m 25s.This is a video-campaign judgment, not a machine-readable robot validator result.

Qwen 3.8 Flash · eiffel-tower-drawing presentation recording poster — Campaign evidence
Qwen 3.8 Flash
Open result
Displayed cost
$0.1320
Wall clock
19m 51.2s
Status
campaign evidence not harness ledger
Reasoning
Max
**Validation scope.** No machine-readable validation outcome supplied
**Material caveat.** Campaign-evidence record; this attempt was not generated from the public Harness ledger.
Clip metric vs result-receipt scope
This permanent campaign-evidence record uses the handoff’s exact displayed cost and wall clock; it is not a generated Harness ledger.
Public record identity: Qwen 3.8 Flash · Qwen 3.8 Flash · Unavailable · Unavailable

DeepSeek V4.1 Flash · eiffel-tower-drawing presentation recording poster — Campaign evidence
DeepSeek V4.1 Flash
Open result
Displayed cost
$0.0841
Wall clock
18m 31.4s
Status
campaign evidence not harness ledger
Reasoning
Max
**Validation scope.** No machine-readable validation outcome supplied
**Material caveat.** Campaign-evidence record; this attempt was not generated from the public Harness ledger.
Clip metric vs result-receipt scope
This permanent campaign-evidence record uses the handoff’s exact displayed cost and wall clock; it is not a generated Harness ledger.
Public record identity: DeepSeek V4.1 Flash · DeepSeek V4.1 Flash · Unavailable · Unavailable

GLM 5.3 Flash · eiffel-tower-drawing presentation recording poster — Campaign evidence
GLM 5.3 Flash
Open result
Displayed cost
$0.1169
Wall clock
10m 28.2s
Status
campaign evidence not harness ledger
Reasoning
Max
**Validation scope.** No machine-readable validation outcome supplied
**Material caveat.** Campaign-evidence record; this attempt was not generated from the public Harness ledger.
Clip metric vs result-receipt scope
This permanent campaign-evidence record uses the handoff’s exact displayed cost and wall clock; it is not a generated Harness ledger.
Public record identity: GLM 5.3 Flash · GLM 5.3 Flash · Unavailable · Unavailable

Gemini 3.8 Flash · eiffel-tower-drawing presentation recording poster — Campaign evidence
Gemini 3.8 Flash
Open result
Displayed cost
$1.36
Wall clock
8m 25s
Status
campaign evidence not harness ledger
Reasoning
High
**Validation scope.** Editorial winner; not a machine-validated score
**Material caveat.** Campaign-evidence record; this attempt was not generated from the public Harness ledger.
Clip metric vs result-receipt scope
This permanent campaign-evidence record uses the handoff’s exact displayed cost and wall clock; it is not a generated Harness ledger.
Public record identity: Gemini 3.8 Flash · Gemini 3.8 Flash · Unavailable · Unavailable

Fixture 14 · episode 06:56

Apple-stem manipulation

Apple Stem · Apple-stem manipulation · Apple Stem Grasp

Editorial observationGLM was the only pass. Gemini lifted the apple using disallowed finger surfaces and therefore did not pass.The pass labels are campaign judgments; no machine-readable robot validator was supplied.

Qwen 3.8 Flash · apple-stem presentation recording poster — Campaign evidence
Qwen 3.8 Flash
Open result
Displayed cost
$0.2397
Wall clock
1h 28m 21.1s
Status
campaign evidence not harness ledger
Reasoning
Max
**Validation scope.** Campaign judgment: did not pass
**Material caveat.** Campaign-evidence record; this attempt was not generated from the public Harness ledger.
Clip metric vs result-receipt scope
This permanent campaign-evidence record uses the handoff’s exact displayed cost and wall clock; it is not a generated Harness ledger.
Public record identity: Qwen 3.8 Flash · Qwen 3.8 Flash · Unavailable · Unavailable

DeepSeek V4.1 Flash · apple-stem presentation recording poster — Campaign evidence
DeepSeek V4.1 Flash
Open result
Displayed cost
$0.0992
Wall clock
1h 28m 20.2s
Status
campaign evidence not harness ledger
Reasoning
Max
**Validation scope.** Campaign judgment: did not pass
**Material caveat.** Campaign-evidence record; this attempt was not generated from the public Harness ledger.
Clip metric vs result-receipt scope
This permanent campaign-evidence record uses the handoff’s exact displayed cost and wall clock; it is not a generated Harness ledger.
Public record identity: DeepSeek V4.1 Flash · DeepSeek V4.1 Flash · Unavailable · Unavailable

GLM 5.3 Flash · apple-stem presentation recording poster — Campaign evidence
GLM 5.3 Flash
Open result
Displayed cost
$0.5583
Wall clock
1h 29m 33.2s
Status
campaign evidence not harness ledger
Reasoning
Max
**Validation scope.** Campaign judgment: pass
**Material caveat.** Campaign-evidence record; this attempt was not generated from the public Harness ledger.
Clip metric vs result-receipt scope
This permanent campaign-evidence record uses the handoff’s exact displayed cost and wall clock; it is not a generated Harness ledger.
Public record identity: GLM 5.3 Flash · GLM 5.3 Flash · Unavailable · Unavailable

Gemini 3.8 Flash · apple-stem presentation recording poster — Campaign evidence
Gemini 3.8 Flash
Open result
Displayed cost
$2.69
Wall clock
1h 9m 1s
Status
campaign evidence not harness ledger
Reasoning
High
**Validation scope.** Campaign judgment: did not pass
**Material caveat.** Campaign-evidence record; this attempt was not generated from the public Harness ledger.
Clip metric vs result-receipt scope
This permanent campaign-evidence record uses the handoff’s exact displayed cost and wall clock; it is not a generated Harness ledger.
Public record identity: Gemini 3.8 Flash · Gemini 3.8 Flash · Unavailable · Unavailable

Overall verdict

Capability moved with the task.

These cheaper Flash models did not show consistent general-purpose performance. No single fixture preference supports a universal leaderboard.

QwenGames and 3D scenes

Several of the strongest game, arena, voxel, and space-flight outputs.

DeepSeekGraphics and CAD delivery

Among the strongest graphics models and the better submitted CAD result, with material CAD defects.

GLMSpecific robot success

The only Apple Stem pass and a single-turn Rubik’s Cube pass.

GeminiRobot speed and drawing

Fastest on the single-turn cube and the editorial Eiffel drawing winner.

LimitNo three-move completion

Only GLM and Gemini advanced; neither completed the extension.

Method disclosure

Receipts outrank narration.

Every clip is hash-bound to the supplied test/model mapping. Clip costs and wall clocks are shown verbatim; reused run receipts keep their own stronger provenance and documented conflicts.

All 54 supplied attempts resolve to unique permanent result pages.
Thirty-five existing public records are reused without duplication; 19 records are explicitly campaign evidence.
Qwen and DeepSeek have no fabricated three-move cube runs because they did not advance.
Robot pass/fail labels remain video-campaign judgments where no machine-readable validator exists.
Gemini’s operator-created CAD assembly and presentation remain post-run work, never scored model completion.
No aggregate score, universal winner, or campaign-specific Builder package has been fabricated.

Inspect all public results Explore benchmark projects and reproduction paths Watch the full episode on YouTube

Previous campaign

Launch 004 remains part of the record.

Revisit Claude Opus 5 across nine game-building tests, including its original video, results, and Shrine reproduction path.

Open Launch 004 Watch Launch 005 on YouTube