OpenAI

Tier B · Specialist or second-choice models.

GPT-6 Luna

B-tier option for bounded builds, with ten published visual runs and a documented context caveat.

Try the active visual tests

Open the generated scene to test it, or compare that prompt with another model.

10 runs

City Scroll Journey

Run date: 2026-09-23

Specialist test · evaluated separately

Generation provenance: unrecorded

Cinematic scroll journey testing AI-generated scene continuity, scroll-scrubbed camera motion, and art-directed landing craft.

Ember Glider

Run date: 2026-09-23

Generation provenance: unrecorded

Superbench 2.1 returning brief · original 1.x run protocol

Sunset gliding journey testing flight energy management, checkpoint flow, and atmospheric scene craft.

Helm's Deep

Run date: 2026-09-23

Generation provenance: unrecorded

Fortress siege scene testing scale, lighting, architecture, and cinematic atmosphere.

Hogwarts Broom Flight Simulator

Run date: 2026-09-23

Generation provenance: unrecorded

Broom-flight scene testing depth, motion cues, castle scale, and fantasy mood.

Low-Poly Tower Defense

Run date: 2026-09-23

Generation provenance: unrecorded

Superbench 2.1 returning brief · original 1.x run protocol

Diorama tower defense testing economy balance, wave design, placement rules, and combat readability.

Mechanical Watch Simulator

Run date: 2026-09-23

Generation provenance: unrecorded

Interactive watch movement testing mechanical legibility, accurate relative motion, and real-time 3D controls.

Neon Drift

Run date: 2026-09-23

Generation provenance: unrecorded

Synthwave time-trial racing testing drift physics, lap timing, ghost replay, and unlock progression.

Starfall Arena

Run date: 2026-09-23

Generation provenance: unrecorded

Neon arena survival testing wave escalation, upgrade builds, particle feedback, and boss design.

Stormwind Trebuchet Simulator

Run date: 2026-09-23

Generation provenance: unrecorded

Counterweight siege simulation testing coupled mechanics, trajectory prediction, projectile cameras, and interactive tuning.

Yingzao Fashi Assembly

Run date: 2026-09-23

Specialist test · evaluated separately

Generation provenance: unrecorded

Timber assembly scene testing structure, joinery, construction order, and material clarity.

Archived model versions

The latest GPT-6 Luna runs are shown above. Earlier versions are preserved separately and are not counted in this current run set.

Editorial verdict

Where this model fits

Back to all benchmarks →

GPT-6 Luna has ten published Superbash visual runs. Eight used fresh agents; two continued unrelated prior tasks because fresh subagent slots were unavailable. Each assignment received only its canonical prompt as creative input, and the methodology records the context difference. B tier is an editorial placement for bounded work with review, not a numeric score.

Best for

  • Small prototypes
  • Bounded UI implementation
  • Iterations with a clear reviewer

Watch out

  • Two runs used sequential continuations with unrelated prior-task context
  • Shared filesystem separation was instruction-based, not independently attested
  • Artifact validation is not a qualitative score
  • Public API pricing and limits were not assessed
Ranking rationale and evidence
  1. Ten active visual artifacts are published with manifests and run-specific checks.
  2. The methodology identifies eight fresh agents and two continuations, so readers can interpret the run set with its context limits.
  3. B tier routes Luna to narrower tasks and stronger review; no universal quality or speed ranking is inferred from the local runs.

Evidence

2026-11-18

Superbash editorial model ranking

Takeaway: GPT-6 Luna is currently placed in Tier B.

The November 2026 editorial roster places GPT-6 Luna at rank 12.

Open source →
2026-09-23

Superbash visual benchmark suite: GPT-6 Luna

Takeaway: Ten Luna runs are published, with eight fresh agents and two documented continuations.

The methodology records the prompt-only assignments, continuation context, and validation limits.

Open source →

Local benchmark run

Prompt-only runs on the active visual suite

Superbash Learn sourceSeptember 23, 2026Run methodology and validation records →
Sources, pricing details and methodology

Run identity

Model specifications

Codex platform ↗
Status
Available in this Codex session
Codex model ID
gpt-6-luna
Run date
September 23, 2026
Access route
Selected in Codex for the September 23, 2026 prompt-only visual runs. Public API availability, pricing, and context limits were not assessed.

Benchmark sources

All eight original 1.x core and two specialist prompts were assigned to GPT-6 Luna. Eight used fresh subagents; two used separate prompt-only continuations after the platform agent-thread cap. No reference implementations, screenshots, or other model outputs were provided as creative material.

GPT-6 Luna prompt-only visual benchmark run

These are inspectable local artifacts, not provider benchmark scores or an aggregate rating. Two continuation assignments retained unrelated prior-task conversation context; the methodology record identifies them. Individual validation reports record browser checks and limitations.

Evaluation settings

    Local visual-run methodology and validation records

    Edition coverage

    Recorded means a published artifact, not a behavioral pass. Historical returning results retain their original protocol.

    Superbench 2.1: 2/6 recorded · 0 pending · 4 not run
    • 15 Ember Glider · repository-run · Recorded
    • 14 Low-Poly Tower Defense · repository-run · Recorded
    • 21 Skyhook Cargo · single-prompt · Not run — Historical returning-brief artifacts only; the four new 2.1 contracts have not been published for this model version.
    • 22 Concourse Divide · single-prompt · Not run — Historical returning-brief artifacts only; the four new 2.1 contracts have not been published for this model version.
    • 23 Lightkeeper's Atelier · single-prompt · Not run — Historical returning-brief artifacts only; the four new 2.1 contracts have not been published for this model version.
    • 24 Wrecklight Salvage · single-prompt · Not run — Historical returning-brief artifacts only; the four new 2.1 contracts have not been published for this model version.