DOOMBENCH.COM
OpenAI 🇺🇸 · GPT-6

GPT-6 Luna

OpenAI's closed GPT-6 Luna reasoning model is available in the API, ChatGPT Work and Codex, with desktop access for Free and Go users. It targets high-volume tasks while the company reports stronger coding and computer-use performance than its prior Luna tier at lower cost.

DOOM SCORE84.0out of 100model risk profile, not the overall index
0 comments · 0 votesOpen discussion

Public discussion is readable by everyone. Sign in to comment, reply, or vote.

No comments yet. Start the discussion.

Send detailed or private feedback →
CURRENT ASSESSMENT · REVISION 1

Why this model scores 84.0

Officially released, inexpensive, widely accessible Luna has substantial coding and tool-use capability, but remains below Sol and Astra on breadth. Its scale of deployment increases exposure and misuse opportunity; stronger reported alignment moderates control difficulty. Benchmarks are mainly vendor-reported.

Capability89
Autonomy86
Deployment100
Misuse potential79
Control difficulty66
WATCH AND EXPLORE

Reviewed videos from Artificial Intelligence Videos. These links add context and never change a DoomBench score.

Thumbnail for GPT-6 Luna First Test - Is OpenAI’s CHEAPEST Model Actually Good?
25:00

GPT-6 Luna First Test - Is OpenAI’s CHEAPEST Model Actually Good?

Bijan Bowen tests GPT-6 Luna on games, a browser desktop, a watch website and a robot arm, weighing basic results against its reported low cost.

The dedicated GPT-6 Luna test covers coding results, visual failures and an unsuccessful robot-arm attempt, directly illustrating this exact model profile; anecdotal demonstrations are not independent benchmark evidence.
MODEL-ATTRIBUTED EVIDENCE

News tied to GPT-6 Luna

The model score of 84.0 rates this model's risk profile. The overall Doom Index of 66.8 measures the complete temporally weighted evidence record. These values answer different questions.

NET MODEL-ATTRIBUTED INDEX CONTRIBUTION+0.07

Each article's current Doom Index contribution is divided equally among the exact models named on that article. This prevents multi-model evidence from being claimed in full on several model pages. Model risk scores use a bounded temporal offset around their technical profile, but never feed back into the overall index.

Capability TOWARD

OpenAI releases GPT-6 Sol and Luna for agentic work at lower cost

OpenAI released GPT-6 Sol and GPT-6 Luna in its API, ChatGPT Work and Codex. The company reports stronger long-horizon coding, professional-work and computer-use evaluations than their GPT-5.6 predecessors while halving listed API prices. The models extend GPT-6-class agent capability to more routine and high-volume workflows, though benchmark results do not establish real-world loss of control.

Full item contribution
+0.11
GPT-6 Luna equal share
+0.05
Read assessment →
AUDIT TRAIL

Model score history

  1. R1
    Doom Score 84.0

    New exact GPT-6 Luna version verified in OpenAI's dated release and API catalogue.

    23 Sept 2026

Share this page

DoomBench social sharing card for GPT-6 Luna.

  1. GPT-6 Luna by OpenAI has a DoomBench model risk score of 84.0 out of 100, based on five transparent version-specific dimensions rather than the overall index.

  2. GPT-6 Luna's highest current DoomBench dimension is deployment at 100.0 out of 100; the profile publishes every component score and its editorial rationale.

  3. DoomBench links 2 source-backed evidence items to GPT-6 Luna, while keeping the model's risk profile separate from each item's contribution to the live Doom Index.

    https://www.doombench.com/models/openai-gpt-6-luna