OpenAILanguage modelActive

GPT-6 Luna

GPT-6 Luna is in the upper middle: Rank 35 of 92 in the overall index. Through the API it is cheap.

IntelligenceRank 35 of 92

96.3Overall indexhigher is better

SpeedRank 40 of 100

101Tokens per secondhigher is better

PriceRank 8 of 95

Input $0.10 · Output $0.50

$0.20per 1M tokenslower is better

ContextRank 2 of 102

1.05Mtokens at once≈ 1,580 A4 pages

PopularityRank 30 of 785

60out of 100+8 vs. last month

based on search, usage and attention

Summary

It is cheap: $0.20 per 1M tokens (input and output blended); a typical model costs $1.55. At 101 tokens per second it writes faster than a typical model (87).

GPT-6 Luna improves multi-step work compared with GPT-5.6 Luna. In OpenAI’s AutomationBench comparison at high reasoning effort, it scores 5.4 percentage points higher at a 58% lower cost per task. The evaluation covers workflows across business applications.

Luna remains the smallest and lowest-cost GPT-6 tier. GPT-6.1 Sol covers the more capable Sol line.

GPT-6 Luna benchmarks

Overall index (Gradually Meta-Score v4)

20 of 92 models

4 models in this view are measured at a lower reasoning effort than max in the overall index. Dashed column: lower reasoning effort than the top.

Claude Opus 5.5max126.1

Gemini 4 Carbonhigh125.8

Claude Fable 5.1max123.8

Claude Sonnet 5.5max122.8

Gemini 4 Argonhigh122.3

GPT-6 Astramax121.3

GPT-6.1 Solmax119.2

Muse Wonder Fullmax118.3

Claude Opus 5max117.2

Muse Wondermax113.4

Kimi K3.1 Promax112.9

Claude Fable 5max112.5

GPT-6 Solmax110.5

GPT-5.6 Solmax108.7

Muse Spark 1.3max107.4

Gemini 3.7 Flashhigh107.1

DeepSeek V4.1 Flash 1017max106.6

Gemini 3.8 Flashhigh106.5

GPT-6.1 Terramax105.3

GPT-6 Lunamax96.3

DeepSeek V4.1 Flash 1017

Gemini 3.8 Flash

GPT-6.1 Terra

GPT-6 Luna

GPT-6 Luna by reasoning effort

Points on the overall index scale. The value at max is GPT-6 Luna's overall index; the other efforts derive from the same Artificial Analysis test series. Missing efforts are never estimated.

GPT-6 Lunanone 59.4 · low 65.2 · medium 82.4 · high 87.1 · xhigh 90.6 · max 96.3DeepSeek-V4.1-Flashnone 84.7 · max 99.7GLM-5.3low 89.4 · max 100.7Filled dot: effort of the overall index. Shading: 95% range.

Source: Meta-Score v4; tiers: Artificial Analysis

Speed

20 of 100 models

Mercury 2.5576tokens/s

Gemini 3.1 Flash-Lite336tokens/s

Gemini 3.5 Flash-Lite299tokens/s

Nemotron 3.5 Lightning245tokens/s

Claude Haiku 5.5236tokens/s

o3-mini230tokens/s

Ling 3.0 Flash199tokens/s

DeepSeek-V3 0324195tokens/s

GPT-5.4 mini191tokens/s

Gemini 3 Flash Preview191tokens/s

Gemini 3.5 Flash184tokens/s

Gemini 3.6 Flash182tokens/s

Command A Plus177tokens/s

DeepSeek-V4.1-Flash176tokens/s

DeepSeek V4.1 Flash 1017172tokens/s

Qwen 3.7 Max165tokens/s

GPT-5.4 nano153tokens/s

GPT OSS 20B152tokens/s

GPT OSS 120B151tokens/s

GPT-6 Luna101tokens/s

Mercury 2.5

Gemini 3.1 Flash-Lite

Gemini 3.5 Flash-Lite

Nemotron 3.5 Lightning

Gemini 3 Flash Preview

DeepSeek V4.1 Flash 1017

GPT-6 Luna

Output tokens per second, measured by Artificial Analysis

Intelligence vs. price

most attractive area: stronger and cheaper than half of the models shownbest intelligence for the price

GPT-6 Luna by task area

Points, 100 = reference group average · higher is better

MathematicsRank 9 of 47105.5

AgentsRank 32 of 9899.5

FinanceRank 38 of 10498.3

ReasoningRank 32 of 7297.8

CodingRank 35 of 10997.5

HealthRank 28 of 7295.2

GeneralRank 31 of 6594.4

MultimodalRank 20 of 4792.0

LegalRank 41 of 7290.7

KnowledgeRank 19 of 3289.2

Text Arena

How people rate the answers in blind comparisons, Elo points with style control · higher is better

20 of 40 models

Gemini 4 Argon1,520

Claude Opus 5.51,509

Muse Wonder1,507

Claude Fable 5.11,504

Claude Fable 51,504

Gemini 3.8 Flash1,496

Muse Spark 1.31,495

Claude Opus 51,491

Gemini 3.7 Flash1,490

Kimi K31,488

Gemini 3.1 Pro Preview1,487

GPT-5.6 Sol1,485

Gemini 3.6 Flash1,483

GPT-6.1 Sol1,481

GLM-5.31,481

Gemini 3.5 Flash1,478

Claude Sonnet 5.51,477

GLM-5.3-Flash1,477

DeepSeek-V4.1-Flash1,476

GPT-6 Luna1,441

Gemini 3.1 Pro Preview

GPT-6 Luna

as of October 30, 2026 · Source: LMArena

WebDev Arena

How people rate generated websites and web apps in blind comparisons, Elo points · higher is better

20 of 43 models

Claude Opus 5.51,810

GPT-6 Astra1,773

Claude Sonnet 5.51,770

GPT-6.1 Sol1,751

Claude Fable 5.11,746

Muse Wonder Full1,725

Claude Opus 51,691

Gemini 4 Argon1,679

GPT-6 Sol1,678

Qwen 3.8 Max 09021,676

DeepSeek V4.1 Flash 10171,665

Kimi K31,656

Muse Spark 1.31,656

Muse Wonder1,652

Grok 4.71,637

Claude Fable 51,624

GLM-5.31,623

DeepSeek-V4.1-Flash1,619

Grok 4.61,616

GPT-6 Luna1,579

DeepSeek V4.1 Flash 1017

GPT-6 Luna

as of October 30, 2026 · Source: LMArena

Terminal-Bench 4.0

Share of tasks the model solves as an agent in the terminal, in percent · higher is better

20 of 38 models

Claude Opus 5.565.2 %

Claude Sonnet 5.564.1 %

Kimi K3.1 Pro63.6 %

Gemini 4 Carbon63.1 %

GPT-6 Astra59.6 %

Claude Fable 5.158.1 %

Muse Wonder Full58.1 %

Gemini 4 Argon57.6 %

GPT-6.1 Sol55.1 %

Claude Opus 553.5 %

GPT-6.1 Terra49.5 %

Muse Wonder49.5 %

GPT-6 Sol44.4 %

Claude Fable 541.4 %

GLM-5.338.9 %

GPT-5.6 Sol37.9 %

GLM-5.3-Flash37.4 %

Claude Haiku 5.536.4 %

Qwen 3.8 Max 090234.3 %

GPT-6 Luna13.6 %

GPT-6 Luna

as of November 7, 2026 · Source: Vals AI

Vibe Code Bench

Share of apps the model builds correctly from a description, in percent · higher is better

20 of 51 models

Claude Sonnet 5.592.4 %

Gemini 4 Argon91.9 %

Claude Haiku 5.590.4 %

Claude Fable 590.4 %

Claude Opus 5.590.3 %

Claude Fable 5.190.3 %

GPT-6 Astra89.6 %

Gemini 4 Carbon89.1 %

GPT-6.1 Sol88.9 %

Claude Opus 588.4 %

GPT-6 Sol87.8 %

DeepSeek V4.1 Flash 101787.7 %

Grok 4.786.2 %

Muse Wonder Full86.0 %

Muse Spark 1.385.9 %

Kimi K385.0 %

Kimi K3.1 Pro84.9 %

DeepSeek-V4.1-Flash84.7 %

GPT-6.1 Terra84.0 %

GPT-6 Luna81.6 %

DeepSeek V4.1 Flash 1017

GPT-6 Luna

as of November 7, 2026 · Source: Vals AI

Show all 1,076 published benchmark results

Published benchmarks

General tasks

BioMysteryBench

Sub-task	GPT-6 Luna (This model)	GPT-6.1 Sol	GPT-6 Astra	GPT-6 Sol	GPT-5.6 Sol	GPT-5.6 Luna
Hard	17.65%	39.22% (best value in this row)	37.26%	35.29%	29.41%	27.45%
Human-solvable	71.69%	89.04% (best value in this row)	89.04% (best value in this row)	84.02%	80.82%	69.41%
Overall	61.48%	79.63% (best value in this row)	79.26%	74.82%	71.11%	61.48%

higher is betterAs of November 7, 2026Source: Vals AIbars scaled within each row, not across rows

Methodology and context

Metric: accuracy

Finance

ClaimBench

Sub-task	GPT-6 Luna (This model)	GPT-6.1 Sol
Direct support	100% (best value in this row)	100% (best value in this row)
Missing evidence	100% (best value in this row)	100% (best value in this row)
Non-supporting citation	100% (best value in this row)	100% (best value in this row)
Numeric discrepancy	100% (best value in this row)	100% (best value in this row)
Overall	99.18%	100% (best value in this row)
Subtle contradiction	95.92%	100% (best value in this row)

higher is betterAs of November 7, 2026Source: Vals AIbars scaled within each row, not across rows

Methodology and context

Metric: accuracy

Recorded measurement details

BioMysteryBench: Hard (Show measurement, test conditions, and source)

Source value: 17.647
Score: 17.647
Metric: accuracy
Unit: %
Benchmark version: 1
Category: general
Task: human_difficult
Task label: Hard
Direction: Higher is better
Comparison cohort: vals-biomysterybench:1:human_difficult
Source type: independent-evaluator
Evaluator: Vals AI
Status: active
Retrieved at: 2026-11-07
Methodology: Source model ID: openai/gpt-6-luna. Source model name: GPT-6 Luna. Provider reported by source: OpenAI. Evaluator: Vals AI. Dataset type: public. Leaderboard date: 2026-11-05. Harness: Terminus 2. Reasoning effort: max. Maximum output tokens: 128000. Latency in seconds: 566.382. Standard error: 3.396. Cost per task in USD: 0.058804. Source ID: vals-ai. Evidence capture ID: vals-biomysterybench-2026-11-07-c2992129e13e. Source payload hash: c2992129e13e1455796b51fe354f18fa049ad9dcfe02fd66196230a60aeea9df. Parser version: vals-astro-v4. Imported at: 2026-11-07
Comparison methodology: Harness: Terminus 2. Reasoning effort: max. Maximum output tokens: 128000

Source Vals AI

BioMysteryBench: Human-solvable (Show measurement, test conditions, and source)

Source value: 71.689
Score: 71.689
Metric: accuracy
Unit: %
Benchmark version: 1
Category: general
Task: human_solvable
Task label: Human-solvable
Direction: Higher is better
Comparison cohort: vals-biomysterybench:1:human_solvable
Source type: independent-evaluator
Evaluator: Vals AI
Status: active
Retrieved at: 2026-11-07
Methodology: Source model ID: openai/gpt-6-luna. Source model name: GPT-6 Luna. Provider reported by source: OpenAI. Evaluator: Vals AI. Dataset type: public. Leaderboard date: 2026-11-05. Harness: Terminus 2. Reasoning effort: max. Maximum output tokens: 128000. Latency in seconds: 513.89. Standard error: 2.416. Cost per task in USD: 0.051939. Source ID: vals-ai. Evidence capture ID: vals-biomysterybench-2026-11-07-c2992129e13e. Source payload hash: c2992129e13e1455796b51fe354f18fa049ad9dcfe02fd66196230a60aeea9df. Parser version: vals-astro-v4. Imported at: 2026-11-07
Comparison methodology: Harness: Terminus 2. Reasoning effort: max. Maximum output tokens: 128000

Source Vals AI

BioMysteryBench: Overall (Show measurement, test conditions, and source)

Source value: 61.481
Score: 61.481
Metric: accuracy
Unit: %
Benchmark version: 1
Category: general
Task: overall
Task label: Overall
Direction: Higher is better
Comparison cohort: vals-biomysterybench:1:overall
Source type: independent-evaluator
Evaluator: Vals AI
Status: active
Retrieved at: 2026-11-07
Methodology: Source model ID: openai/gpt-6-luna. Source model name: GPT-6 Luna. Provider reported by source: OpenAI. Evaluator: Vals AI. Dataset type: public. Leaderboard date: 2026-11-05. Harness: Terminus 2. Reasoning effort: max. Maximum output tokens: 128000. Latency in seconds: 523.805. Standard error: 2.593. Cost per task in USD: 0.053236. Source ID: vals-ai. Evidence capture ID: vals-biomysterybench-2026-11-07-c2992129e13e. Source payload hash: c2992129e13e1455796b51fe354f18fa049ad9dcfe02fd66196230a60aeea9df. Parser version: vals-astro-v4. Imported at: 2026-11-07
Comparison methodology: Harness: Terminus 2. Reasoning effort: max. Maximum output tokens: 128000

Source Vals AI

ClaimBench: Direct support (Show measurement, test conditions, and source)

Source value: 100
Score: 100
Metric: accuracy
Unit: %
Benchmark version: 1
Category: finance
Task: direct_support
Task label: Direct support
Direction: Higher is better
Comparison cohort: vals-claimbench:1:direct_support
Source type: independent-evaluator
Evaluator: Vals AI
Status: active
Retrieved at: 2026-11-07
Methodology: Source model ID: openai/gpt-6-luna. Source model name: GPT-6 Luna. Provider reported by source: OpenAI. Evaluator: Vals AI. Dataset type: public. Leaderboard date: 2026-10-23. Reasoning effort: max. Maximum output tokens: 128000. Latency in seconds: 6.911378. Standard error: 0. Cost per task in USD: 0.00132. Source ID: vals-ai. Evidence capture ID: vals-claimbench-2026-11-07-6aa89f0db85e. Source payload hash: 6aa89f0db85e2bdc765ff8d942de0d491dd37f2f2e635bb4a95b0f92746fc6d3. Parser version: vals-astro-v4. Imported at: 2026-11-07
Comparison methodology: Reasoning effort: max. Maximum output tokens: 128000

Source Vals AI

ClaimBench: Missing evidence (Show measurement, test conditions, and source)

Source value: 100
Score: 100
Metric: accuracy
Unit: %
Benchmark version: 1
Category: finance
Task: missing_evidence
Task label: Missing evidence
Direction: Higher is better
Comparison cohort: vals-claimbench:1:missing_evidence
Source type: independent-evaluator
Evaluator: Vals AI
Status: active
Retrieved at: 2026-11-07
Methodology: Source model ID: openai/gpt-6-luna. Source model name: GPT-6 Luna. Provider reported by source: OpenAI. Evaluator: Vals AI. Dataset type: public. Leaderboard date: 2026-10-23. Reasoning effort: max. Maximum output tokens: 128000. Latency in seconds: 7.086452. Standard error: 0. Cost per task in USD: 0.001385. Source ID: vals-ai. Evidence capture ID: vals-claimbench-2026-11-07-6aa89f0db85e. Source payload hash: 6aa89f0db85e2bdc765ff8d942de0d491dd37f2f2e635bb4a95b0f92746fc6d3. Parser version: vals-astro-v4. Imported at: 2026-11-07
Comparison methodology: Reasoning effort: max. Maximum output tokens: 128000

Source Vals AI

ClaimBench: Non-supporting citation (Show measurement, test conditions, and source)

Source value: 100
Score: 100
Metric: accuracy
Unit: %
Benchmark version: 1
Category: finance
Task: non_supporting_citation
Task label: Non-supporting citation
Direction: Higher is better
Comparison cohort: vals-claimbench:1:non_supporting_citation
Source type: independent-evaluator
Evaluator: Vals AI
Status: active
Retrieved at: 2026-11-07
Methodology: Source model ID: openai/gpt-6-luna. Source model name: GPT-6 Luna. Provider reported by source: OpenAI. Evaluator: Vals AI. Dataset type: public. Leaderboard date: 2026-10-23. Reasoning effort: max. Maximum output tokens: 128000. Latency in seconds: 7.176003. Standard error: 0. Cost per task in USD: 0.001385. Source ID: vals-ai. Evidence capture ID: vals-claimbench-2026-11-07-6aa89f0db85e. Source payload hash: 6aa89f0db85e2bdc765ff8d942de0d491dd37f2f2e635bb4a95b0f92746fc6d3. Parser version: vals-astro-v4. Imported at: 2026-11-07
Comparison methodology: Reasoning effort: max. Maximum output tokens: 128000

Source Vals AI

ClaimBench: Numeric discrepancy (Show measurement, test conditions, and source)

Source value: 100
Score: 100
Metric: accuracy
Unit: %
Benchmark version: 1
Category: finance
Task: numeric_discrepancy
Task label: Numeric discrepancy
Direction: Higher is better
Comparison cohort: vals-claimbench:1:numeric_discrepancy
Source type: independent-evaluator
Evaluator: Vals AI
Status: active
Retrieved at: 2026-11-07
Methodology: Source model ID: openai/gpt-6-luna. Source model name: GPT-6 Luna. Provider reported by source: OpenAI. Evaluator: Vals AI. Dataset type: public. Leaderboard date: 2026-10-23. Reasoning effort: max. Maximum output tokens: 128000. Latency in seconds: 7.372316. Standard error: 0. Cost per task in USD: 0.001414. Source ID: vals-ai. Evidence capture ID: vals-claimbench-2026-11-07-6aa89f0db85e. Source payload hash: 6aa89f0db85e2bdc765ff8d942de0d491dd37f2f2e635bb4a95b0f92746fc6d3. Parser version: vals-astro-v4. Imported at: 2026-11-07
Comparison methodology: Reasoning effort: max. Maximum output tokens: 128000

Source Vals AI

ClaimBench: Overall (Show measurement, test conditions, and source)

Source value: 99.184
Score: 99.184
Metric: accuracy
Unit: %
Benchmark version: 1
Category: finance
Task: overall
Task label: Overall
Direction: Higher is better
Comparison cohort: vals-claimbench:1:overall
Source type: independent-evaluator
Evaluator: Vals AI
Status: active
Retrieved at: 2026-11-07
Methodology: Source model ID: openai/gpt-6-luna. Source model name: GPT-6 Luna. Provider reported by source: OpenAI. Evaluator: Vals AI. Dataset type: public. Leaderboard date: 2026-10-23. Reasoning effort: max. Maximum output tokens: 128000. Latency in seconds: 7.283. Standard error: 0.203. Cost per task in USD: 0.001372. Source ID: vals-ai. Evidence capture ID: vals-claimbench-2026-11-07-6aa89f0db85e. Source payload hash: 6aa89f0db85e2bdc765ff8d942de0d491dd37f2f2e635bb4a95b0f92746fc6d3. Parser version: vals-astro-v4. Imported at: 2026-11-07
Comparison methodology: Reasoning effort: max. Maximum output tokens: 128000

Source Vals AI

ClaimBench: Subtle contradiction (Show measurement, test conditions, and source)

Source value: 95.918
Score: 95.918
Metric: accuracy
Unit: %
Benchmark version: 1
Category: finance
Task: subtle_contradiction
Task label: Subtle contradiction
Direction: Higher is better
Comparison cohort: vals-claimbench:1:subtle_contradiction
Source type: independent-evaluator
Evaluator: Vals AI
Status: active
Retrieved at: 2026-11-07
Methodology: Source model ID: openai/gpt-6-luna. Source model name: GPT-6 Luna. Provider reported by source: OpenAI. Evaluator: Vals AI. Dataset type: public. Leaderboard date: 2026-10-23. Reasoning effort: max. Maximum output tokens: 128000. Latency in seconds: 7.866128. Standard error: 1.001. Cost per task in USD: 0.001357. Source ID: vals-ai. Evidence capture ID: vals-claimbench-2026-11-07-6aa89f0db85e. Source payload hash: 6aa89f0db85e2bdc765ff8d942de0d491dd37f2f2e635bb4a95b0f92746fc6d3. Parser version: vals-astro-v4. Imported at: 2026-11-07
Comparison methodology: Reasoning effort: max. Maximum output tokens: 128000

Source Vals AI

Page 1 of 89

See for yourself: one prompt, one attempt

We show the latest attempt without fixes; it does not count toward the overall index.

Pelican on a bicycle

Prompt

Generate an SVG of a pelican riding a bicycle

Oct 1, 2026 · 47 s · $0.0022

Test details

Run date: Oct 1, 2026
API provider: openrouter
Requested API model: openai/gpt-6-luna
Reasoning setting: high
Token limit including reasoning: 8,192 tokens
Output tokens: 4,493 tokens
Cost of the attempt: $0.0022
Response time: 46.7 s
API model reported in response: openai/gpt-6-luna
Pinned endpoint: OpenAI | openai/gpt-6-luna-20260922
Temperature: Provider default
Test protocol: community-visual-tests-v1
Technical checks: Passed, not a quality rating

Visually reviewed. This individual test does not establish overall model quality.

Task origin: Simon Willison

SVG short test

One SVG task with the same fixed settings for every model, without hidden thinking.

GPT-6 Luna

Unavailable in this round

Popularity compared

How often the model is searched for, used and mentioned, compared with every model of its category.

Popularity index

The 100 most popular models of this category, 0 to 100, higher is more popular

20 of 100 models

GPT-6 Astra83

Claude Haiku 5.577

GPT-6 Sol73

GPT-573

DeepSeek V4.1 Flash 101769

Kimi K367

MiniCPM-V 4.7 35B-A3B63

GPT-6 Luna60

DeepSeek V4.1 Flash 1017

MiniCPM-V 4.7 35B-A3B

GPT-6.1 Sol

Claude Opus 5

Qwen 3.8 Flash

GPT-6 Luna

As of November 2026 · Source: gradually.ai popularity index

Over time

Weekly values from October 18, 2026 to November 7, 2026

GPT-6 LunaGemini 3.8 Flash 61Muse Spark 1.3 61Claude Sonnet 5.5 60GPT-5.4 mini 59

Method changes in the timeline, a jump at these points is no market event:

1November 1, 2026: Value only with search or usage data, 30-day attention

What is gaining popularity right now, across models, vendors, harnesses and AI subscriptions, is collected in All AI trends.

GPT-6 Luna compared

Compare with another model

First modelSecond model

Predecessor and successor

PredecessorGPT-6 Luna vs. GPT-5.6 LunaPredecessor of GPT-6 Luna

The most popular comparisons

Show all 56 comparisons

Similar models

More models of the same kind as GPT-6 Luna, chosen by model line, developer, size, price and release date.

All specifications

Architecture

Knowledge cutoff: May 18, 2026; Source: Knowledge cutoff

Input and output

Context window: 1,050,000 tokens; Source: Context window
Maximum output: 128,000 tokens; Source: Maximum output
Input: Text, Image; Source: Input
Output: Text; Source: Output
Reasoning mode: none, low, medium, high, xhigh, max; Source: Reasoning mode
Tool calling: Yes; Source: Tool calling

Watermarking and provenance

AI watermarking refers to invisible signals embedded in the output. Visible marks and provenance metadata are listed separately. Unknown means there is insufficient evidence for that particular field.

AI watermarking: Unknown; no source
Watermarking technique: Not known; no source

Licence and access

Weights: Proprietary; Source: Weights
Access: API; Source: Access
Release: Sep 22, 2026; Source: Release

API

API model ID: gpt-6-luna; Source: API model ID

Further details

Notes: Released September 22, 2026 as the lowest-cost GPT-6 tier, described by OpenAI as its most efficient model for focused, high-volume tasks. 1.05 million token context window, 128,000 maximum output tokens, and a May 18, 2026 knowledge cutoff. API price $0.10 / $0.50 per million input and output tokens, half of GPT-5.6 Luna's promotional price. Reasoning effort from none to max, with medium as the default. Official API ID: gpt-6-luna.; Source: Notes

Prices

API pricesPrice historySubscriptions

9 offers, prices in US dollars per 1M tokens, blended from 3 parts input and 1 part output.

Provider	Input	Output	Cache read	Cache write	Blended
OpenAIdirect	$0.10	$0.50	$0.01	$0.125	$0.20
OpenAIabove 272,000 context tokensdirect	$0.20	$0.75	$0.02	$0.25	$0.34
OpenAIFlex	$0.05	$0.25	n/a	n/a	$0.10
Amazon BedrockGlobal	$0.10	$0.50	n/a	n/a	$0.20
Azure	$0.10	$0.50	n/a	n/a	$0.20
OpenAI	$0.10	$0.50	n/a	n/a	$0.20
Amazon BedrockUs-east-1	$0.11	$0.55	n/a	n/a	$0.22
AzureEU	$0.11	$0.55	n/a	n/a	$0.22
AzureUS	$0.11	$0.55	n/a	n/a	$0.22
OpenAIFast	$0.20	$1	n/a	n/a	$0.40

Cache prices come from the direct tariff; OpenRouter does not report them for the other providers. Direct price according to the developer, checked Sep 22, 2026. Provider prices via OpenRouter, as of Nov 7, 2026. Open the LLM price index

2 price changes since tracking began.

InputOutput

Daily observed price in the LLM price index in US dollars per 1M tokens. The index uses a market price from several sources, which can differ from the developer's direct price.

Free to use at OpenAI (ChatGPT Free), more from $8 a month (ChatGPT Go). Through other providers from $1 a month (Command Code Go, with limited credits). 57 plans in Germany.

Plan	Monthly list price	Note
ChatGPT FreeOpenAI	free	developer's own plan
ChatGPT GoOpenAI	$8	developer's own plan
ChatGPT PlusOpenAI	$20	developer's own plan
ChatGPT Business Standard-PlatzOpenAI	$25	developer's own plan, with limited credit, for teams, per seat
ChatGPT Pro $100OpenAI	$100	developer's own plan, with limited credit
ChatGPT Business Premium-PlatzOpenAI	$125	developer's own plan, with limited credit, for teams, per seat
ChatGPT EnterpriseOpenAI	n/a	developer's own plan, for enterprises, per seat
ChatGPT Pro $500OpenAI	$500	developer's own plan, with limited credit
ElevenLabs Free	free	with limited credit
Genspark Free	free	with limited credit
JetBrains AI Free	free	with limited credit
Lovable Free	free	with limited credit
Zed Student	free
Command Code Go	$1	with limited credit
ElevenLabs Starter	$6	with limited credit
Command Code GOAT	$10	with limited credit
GitHub Copilot Pro	$10	with limited credit
Zed Pro	$10
JetBrains AI Pro	$10	with limited credit
GitHub Copilot Business	$19	with limited credit, for teams, per seat
Command Code Pro	$20	with limited credit
Replit Core	$20
ElevenLabs Creator	$22	with limited credit
JetBrains AI Pro	$20	with limited credit, for teams, per seat
Lovable Pro	$25	with limited credit
Genspark Plus	$24.99	with limited credit
Langdock Business Standard	€25	for teams, per seat
JetBrains AI Ultimate	$30	with limited credit
GitHub Copilot Enterprise	$39	with limited credit, for enterprises, per seat
GitHub Copilot Pro+	$39	with limited credit
Genspark for Business Team	$40	with limited credit, for teams, per seat
Lovable Business	$50	with limited credit, for teams
Mindverse Studio Team	€49	for teams, per seat
Higgsfield Plus	€59	with limited credit
Mindverse Studio Pro	€59	per seat
JetBrains AI Ultimate	$60	with limited credit, for teams, per seat
Higgsfield Team	€79	with limited credit, for teams, per seat
ElevenLabs Pro	$99	with limited credit
Command Code Max 10×	$100	with limited credit
GitHub Copilot Max	$100	with limited credit
Replit Pro	$100
Krea Max	$105	with limited credit
Langdock Business Max	€99	for teams, per seat
Higgsfield Ultra	€129	with limited credit
Higgsfield Scale	€169	with limited credit, for teams, per seat
Command Code Max 20×	$200	with limited credit
Krea Business	$200	with limited credit, for teams
Genspark Pro	$249.99	with limited credit
ElevenLabs Scale	$299	with limited credit, for teams
ElevenLabs Business	$990	with limited credit, for teams
ElevenLabs Enterprise	n/a	for enterprises
Genspark for Business Enterprise	n/a	with limited credit, for enterprises, per seat
Higgsfield Enterprise	n/a	for enterprises
JetBrains AI Enterprise	n/a	with limited credit, for enterprises, per seat
Krea Enterprise	n/a	for enterprises
Lovable Enterprise	n/a	for enterprises
Replit Enterprise	n/a	for enterprises

All plans in the AI subscription comparison

Frequently asked questions about GPT-6 Luna

How good is GPT-6 Luna in benchmarks?

Decent: Rank 35 of 92 in the overall index.

Overall index (Gradually Meta-Score v4) from 48 benchmark results · data through Nov 7, 2026

IntelligenceOverall index in points, 100 = reference group average · higher is better

Step 5 Preview97.3

Gemini 3.5 Flash96.9

Gemini 3.6 Flash96.7

GPT-6 LunaRank 3596.3

Grok 4.594.6

GLM-5.3-Flash94.2

GPT-5.6 Luna93.9

What is the Overall index (Gradually Meta-Score v4)?

A point value that combines many independent benchmarks into one number. 100 points is the average of a fixed reference group, higher is better. GPT-6 Luna scores 96.3 points, which is rank 35 of 92.

Own analysis by gradually.ai, method benchmark-meta-score-v4 · for GPT-6 Luna from 48 test series

Which data: Only current, reviewed results from independent evaluators such as Artificial Analysis, Vals AI, LMArena, LiveBench and Epoch AI. Vendor claims and historical or disputed values do not count.
One value per test: For each source and test series the latest retrieval with one fixed setup counts. When a source tests several reasoning levels, the highest documented one applies. A series needs at least 5 models.
Making tests comparable: Percentages are converted so that gaps near 0 and near 100 % count fairly, then every series is put on the same scale. A hard test therefore does not weigh more than an easy one.
Shared strength: A statistical model estimates, from all series at once, one shared capability per model and, per area, how far a model deviates from it. Missing tests are never counted as zero.
Weighting: Core areas (Coding, Reasoning, Mathematics, Knowledge, Agents, General and Multimodal) count 75 % together, the domain areas Finance, Legal and Health 25 %.
The point scale: The value is measured against a fixed reference group of 29 models, much like an IQ score: 100 points is its average and 15 points one standard deviation of the group. The scale is open at both ends and is not a percentage. A model 15 points ahead does better in roughly three of four benchmark comparisons. New models do not shift older values.
Place by score: The place follows the order of the scores among all models with a place, such as "rank 14 of 95". A second calculation checks whether a lead is reliable by recomputing the score 200 times with randomly redrawn test families. If a model is not almost always ahead there, the two models count as level.
Minimum evidence: A value needs at least 8 test series from 2 sources and 3 areas, a place only with enough precision. Missing core areas are flagged as "estimated".
What it is not: Not a test run by gradually.ai and no verdict on every single task. For a specific task the area values and the individual benchmarks matter more.

How much does GPT-6 Luna cost?

Through the API $0.10 per 1M input tokens and $0.50 output. That is cheap.

Prices from Nov 7, 2026

Output price since tracking beganDollars per 1M tokens · lower is better

Sep 22, 2026Nov 7, 2026

First price on Sep 22, 2026: $0.50. Since Sep 24, 2026: $0.50.

What is the knowledge cutoff of GPT-6 Luna?

May 18, 2026, according to OpenAI. That is 4 months before release.

Stated by OpenAI

FebMarAprMayJunJulAugSepKnowledge cutoffMay 18, 2026ReleasedSep 22, 2026GPT-5.6 Luna cutoffFeb 16, 2026

How large is the context window of GPT-6 Luna?

1.05 M tokens. That is how much text GPT-6 Luna reads at once, about 1,580 A4 pages.

Page count as a rule of thumb from Artificial Analysis

Context windowTokens the model reads at once · higher is better

GPT-6 LunaRank 2 of 1021.05M

GPT-5.6 Luna1.05M

Gemini 3.6 Flash1.05M

Gemini 3.5 Flash1.05M

GLM-5.3-Flash1M

Claude Opus 4.71M

Grok 4.5500K

How fast is GPT-6 Luna?

101 tokens per second. That is faster than a typical model at 87.

Measured by Artificial Analysis on Nov 6, 2026. AA 3.0.0, including time to first token, averaged across available workloads and five days.

SpeedTokens per second · higher is better

Gemini 3.5 Flash184

Gemini 3.6 Flash182

Step 5 Preview132

GPT-6 LunaRank 40 of 100101

GPT-5.6 Luna94

Grok 4.571

GLM-5.3-Flash56

Can I run GPT-6 Luna locally?

No. OpenAI does not publish the weights.

Stated by OpenAI

Open weights: no
License: n/a
Parameters: n/a
active per token: n/a

Where can I use GPT-6 Luna?

Free to use at OpenAI. Including other providers that makes 57 plans in Germany, plus the API.

Subscription comparison, prices in Germany incl. VAT

At OpenAI

8 plans, free (ChatGPT Free).

Through other providers

49 plans from 12 providers, from $1 a month (Command Code Go), with limited credits.

Through the API

$0.10 input, $0.50 output per 1M tokens, unchanged since Sep 24, 2026.

Is GPT-6 Luna better than GPT-5.6 Luna?

Not clearly in the overall index: both are level. It leads in Mathematics and Knowledge.

GPT-5.6 Luna as predecessor according to Introducing GPT-6 Sol and Luna, checked Sep 22, 2026

IntelligenceOverall index, 100 points = average · higher is better

96.3

GPT-6 Luna

Rank 35

93.9

GPT-5.6 Luna

Rank 38

SpeedTokens per second · higher is better

101

GPT-6 Luna

fairly fast

GPT-5.6 Luna

fairly fast

PriceDollars per 1M tokens, 3 parts input to 1 part output · lower is better

$0.20

GPT-6 Luna

cheaper

$0.45

GPT-5.6 Luna

2.3 times the price

AreaGPT-6 LunaGPT-5.6 Luna100 points = average, higher is betterResult

Codinglevel13 and 11 tests97.596.0

Reasoninglevel4 and 3 tests97.898.1

MathematicsGPT-6 Luna ahead105.594.9

KnowledgeGPT-6 Luna ahead89.283.9

Agentslevel4 and 6 tests99.597.6

GeneralGPT-6 Luna slightly aheadnot certain · 5 tests each94.490.9

Financelevel5 and 6 tests98.398.2

LegalGPT-5.6 Luna slightly aheadnot certain · 3 and 2 tests90.795.2

Healthlevel2 tests each95.292.4

Multimodallevel2 and 3 tests92.092.2

Sources and data date

Show all 112 sources

Benchmark sources at a glance

Which independent evaluators tested this model. This is coverage, not a quality ranking.

Evaluator	Test series	Composite indices	Recorded active measurements
Artificial Analysis	14	0	86
Vals AI	28	5	291
LMArena	4	0	43
LiveBench	7	1	46
Epoch AI	6	0	13
ARC Prize	0	0	24
Scale AI Evals	0	0	4

Without vendor claims and historical or disputed results; those stay labelled in the single values. 42 measurements are not yet assigned to a test series.

Type	Evidence and data date
Speed	OpenAI, Max effort. AA 3.0.0, including time to first token, averaged across available workloads and five days. Retrieved November 6, 2026. Source: Artificial Analysis · Speed (Artificial Analysis)Speed measurement methodology
Price	Base API prices without caching or batch discounts. Higher context tiers and other rates are listed under Costs. Source: Pricing · Context window data (retrieved September 22, 2026) · GPT-6 Luna API model documentation · Editorial description reviewed October 4, 2026
Input	Documented modalities. Retrieved September 22, 2026. Source: Model documentation · Context window data (retrieved September 22, 2026) · GPT-6 Luna API model documentation · Editorial description reviewed October 4, 2026
Output	Documented modalities. Retrieved September 22, 2026. Source: Model documentation · Context window data (retrieved September 22, 2026) · GPT-6 Luna API model documentation · Editorial description reviewed October 4, 2026
Research status	Research date September 22, 2026. 73 source URLs checked. This documents the inspected sources, not an exhaustive inventory of every publication.
Not found in the inspected sources	parameters-billions Original wording OpenAI does not publish parameter counts.
Unresolved	fresh-agent-arena-leaderboard Note The current Agent Arena leaderboard was reviewed in the rendered web reader, but the raw endpoint returned HTTP 403. No dynamic values were added without a complete captured source body.
Not published	confidence-interval-95 Original wording No per-model 95% interval for this row is published in the paper’s Table 6; no interval has been inferred.
Not published	api-snapshot-id Original wording The paper gives the exact API alias and release date but not a pinned provider snapshot identifier.
Not published	serving-revision-and-test-denominators Note ARC Prize pages do not state per-row N, CI, or immutable served revision. Parallel omits model snapshot and varies reasoning/token limits by provider guidance.
Unresolved	model snapshot, evaluation run date and per-model N Original wording The opened public benchmark article does not give GPT-6 Luna's pinned API snapshot, an evaluation run date for this score table, or a per-model sample count. None is inferred from dataset descriptions.
Not published	PRBench row subset and model-specific run settings Note The current leaderboard page does not identify whether this row uses the full or hard subset, or publish the model-specific generation settings or measurement date.
Additional source	Model metadata (retrieved September 22, 2026) · Introducing GPT-6 Sol and GPT-6 Luna · Editorial description reviewed October 4, 2026
Additional source	Artificial Analysis (retrieved October 19, 2026; November 2, 2026; November 7, 2026)1 2 3 4 5 6 7 8 9
Additional source	Artificial Analysis (retrieved October 19, 2026; October 29, 2026; November 2, 2026; November 7, 2026)1 2
Additional source	Artificial Analysis (retrieved October 19, 2026)1 2
Additional source	Artificial Analysis (retrieved October 19, 2026; November 2, 2026)1 2
Additional source	Artificial Analysis (retrieved November 5, 2026; November 7, 2026)
Additional source	Artificial Analysis (retrieved November 7, 2026)1 2
Additional source	LMArena (retrieved October 30, 2026)1 2 3 4
Additional source	LMArena (retrieved October 28, 2026)1 2
Additional source	LMArena (retrieved October 16, 2026)1 2 3 4
Additional source	Vals AI (retrieved November 7, 2026)1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26
Additional source	Vals AI (retrieved October 16, 2026; November 7, 2026)1 2 3
Additional source	Vals AI (retrieved October 16, 2026; October 31, 2026; November 7, 2026)
Additional source	Vals AI (retrieved October 17, 2026; November 7, 2026)
Additional source	Vals AI (retrieved October 16, 2026; October 30, 2026; October 31, 2026)
Additional source	Vals AI (retrieved October 28, 2026; November 7, 2026)
Additional source	LiveBench (retrieved October 16, 2026)1 (unavailable on November 2, 2026) https://livebench.ai/table_2026_09_02.csv2 (unavailable on November 2, 2026) https://livebench.ai/breakdown_2026_09_02.json
Additional source	Arena AI (retrieved October 17, 2026)
Additional source	PolicyBench (PolicyEngine), official frozen US paper and leaderboard (retrieved October 17, 2026)
Additional source	ARC Prize Verified Leaderboard (retrieved October 17, 2026)
Additional source	Parallel Search Capability Leaderboard (retrieved October 17, 2026)
Additional source	Android Developers, Android Bench leaderboard (retrieved October 17, 2026)
Additional source	VetEval, GPT-6 Luna model page (retrieved October 17, 2026)
Additional source	BANGLAPLAN research team, Constraint-Based Planning and Decision-Making in Bangla (retrieved October 17, 2026)
Additional source	Joonlab GPT-6.1 Sol benchmark, pinned data commit (retrieved October 17, 2026)1 2 3
Additional source	Epoch AI, AI systems are improving rapidly at Slay the Spire and its sequel (retrieved October 17, 2026)
Additional source	VeriContest paper authors (retrieved October 18, 2026)
Additional source	Epoch AI benchmark run dataset snapshot (retrieved October 19, 2026; October 28, 2026)
Additional source	AssemblyAI Voice Standard Bench (retrieved October 19, 2026)
Additional source	Roboflow Vision Evals (retrieved October 19, 2026; October 28, 2026)
Additional source	Agent Arena (retrieved October 20, 2026)
Additional source	VulcanBench Frontier v4, Code quality protocol v3.16 (retrieved October 20, 2026)
Additional source	OpenAI (retrieved October 22, 2026)
Additional source	uhyo, React Proficiency Benchmark (retrieved October 22, 2026)
Additional source	OpenAI Deployment Safety Hub, GPT-6.1 Sol system-card addendum (retrieved October 23, 2026)
Additional source	OpenAI, GPT-6.1 Sol System Card Addendum (retrieved October 23, 2026)
Additional source	LiveBench (retrieved October 28, 2026)1 2
Additional source	Scale AI (retrieved October 28, 2026)
Additional source	Terminal-Bench (retrieved October 28, 2026)
Additional source	DuelLab GameBench 2 model results (retrieved October 28, 2026; November 2, 2026; November 5, 2026)
Additional source	Parallel Search Capability Leaderboard (retrieved October 28, 2026; November 5, 2026)
Additional source	Towards AI ToneBench (retrieved October 28, 2026; October 30, 2026)1 2 3
Additional source	Google on Kaggle (retrieved October 28, 2026)1 2 3 4
Additional source	MathBench.ai (retrieved October 28, 2026)
Additional source	Scale AI HSS (retrieved October 28, 2026)
Additional source	Scale AI DistressBench (retrieved October 28, 2026)
Additional source	Scale AI FORTRESS (retrieved October 28, 2026)
Additional source	Parallel Search Capability Leaderboard (retrieved November 5, 2026)
Additional source	OpenRouter: Popularity index (derived), CC BY 4.0
Additional source	Vercel AI Gateway: Popularity index (derived), CC BY 4.0
Additional source	Ollama: cumulative downloads for the popularity index
Additional source	Replicate: cumulative runs for the popularity index
Additional source	Google Trending Now: limited feed selection for Germany and the US
Additional source	Creative Commons Attribution 4.0