Skip to main content
Van Data Team

text model comparison

GPT-5.6 Luna vs Grok 4.6

GPT-5.6 Luna and Grok 4.6 are both current text models, from OpenAI and xAI. GPT-5.6 Luna is 90% cheaper per 1M input tokens than Grok 4.6, $0.20 vs $2. On Arena Text, GPT-5.6 Luna scores 1452 Elo and Grok 4.6 1454 Elo (Arena, measured October 8, 2026). GPT-5.6 Luna has the larger context window, 1,050,000 vs 500,000 tokens.

Higher Arena Text
Grok 4.61454 Elo against 1452 Elo
Lower price per 1M input tokens
GPT-5.6 Luna$0.20 against $2
Longer context window
GPT-5.6 Luna1.05M tokens against 500K tokens
Newer release
Grok 4.6Aug 2026 against Jul 2026

Data verified October 12, 2026 · Dataset 2026-10-28

Share this

Side by side

GPT-5.6 Luna and Grok 4.6, line by line

Prices in effect today, compared at the same tier wherever both list one. The better value in each row is highlighted.

FactGPT-5.6 LunaGrok 4.6
Price
Price per 1M input tokens$0.20 (better)$2
Price per 1M output tokens$1.20 (better)$6
Independent benchmarks
Arena Text1452 Elo (#34 of 49)1454 Elo (#33 of 49) (better)
GPQA Diamond91.6% (#20 of 40)94% (#9 of 40) (better)
FrontierMath Tiers 1-3 (v2)82.1% (#12 of 40) (better)66% (#24 of 40)
SimpleQA Verified41% (#27 of 36)49.3% (#20 of 36) (better)
Specs
ProviderOpenAIxAI
Release dateJuly 9, 2026August 12, 2026
AccessAPI onlyAPI only
Context window1,050,000 tokens500,000 tokens
Max output128,000 tokensNot reported
Inputtext, imagetext, image
Outputtexttext

Scores:Arena · 2026-10-12Epoch AI · 2026-10-12

Open both in the interactive comparison

Charts

Against every current text model

Where GPT-5.6 Luna and Grok 4.6 land among the current models of their task, on every independent benchmark, on price and on context.

Where each stands

On Arena Text, GPT-5.6 Luna ranks #34 and Grok 4.6 #33 of 49 current models. Each bar is the share of current models it beats: on score, price and context.

GPT-5.6 Luna and Grok 4.6 among the current text models, one row per benchmark, price and context window.

Independent benchmarks

Arena Text

GPT-5.6 Luna:#34of 49

1,452 Elo · ahead of 31%

Grok 4.6:#33of 49

1,454 Elo · ahead of 33%

GPQA Diamond

GPT-5.6 Luna:#20of 40

91.6% · ahead of 51%

Grok 4.6:#9of 40

94% · ahead of 79%

FrontierMath Tiers 1-3 (v2)

GPT-5.6 Luna:#12of 40

82.1% · ahead of 72%

Grok 4.6:#24of 40

66% · ahead of 41%

SimpleQA Verified

GPT-5.6 Luna:#27of 36

41% · ahead of 23%

Grok 4.6:#20of 36

49.3% · ahead of 46%

Price and context

Price per 1M input tokens

GPT-5.6 Luna:#7of 51

$0.20 · cheaper than 84%

Grok 4.6:#28of 51

$2 · cheaper than 24%

Price per 1M output tokens

GPT-5.6 Luna:#10of 51

$1.20 · cheaper than 78%

Grok 4.6:#27of 51

$6 · cheaper than 42%

Context window

GPT-5.6 Luna:#1of 66

1.05M tokens · longer than 86%

Grok 4.6:#54of 66

500K tokens · longer than 15%

Scores fromArena · 2026-10-12Epoch AI · 2026-10-12

Quality against price

Arena Text · measured October 8, 2026

GPT-5.6 Luna scores 1,452 at $0.45; Grok 4.6 scores 1,454 at $3 (blended price per 1M tokens). The line joins the current models no other beats on both price and Arena Text score; hover any point to name it.

Quality against priceArena Text score against price for 41 text models, GPT-5.6 Luna and Grok 4.6 marked.1,2501,3001,3501,4001,4501,5001,550$0.30$1$3$10MOST ATTRACTIVE QUADRANTMedian price $3Median 1,475Blended price per 1M tokens (3 input : 1 output), log scaleArena Text (Elo)Grok 4.6GPT-5.6 Luna
Quality against price: the data
Arena Text score against price for 41 text models, GPT-5.6 Luna and Grok 4.6 marked.
ModelProviderBlended price per 1M tokensArena Text (Elo)
Gemini 4 ArgonGoogle$41,525
Claude Opus 5.5Anthropic$81,507
Claude Fable 5Anthropic$201,504
Claude Fable 5.1Anthropic$201,501
Gemini 3.8 FlashGoogle$1.501,497
Muse Spark 1.3Meta$21,494
Claude Opus 5Anthropic$101,490
Kimi K3Moonshot Kimi$61,488
Gemini 3.7 FlashGoogle$1.501,486
GPT-5.6 SolOpenAI$81,485
GPT-6.1 SolOpenAI$41,484
Qwen3.8-MaxAlibaba Qwen$31,483
GPT-5.5OpenAI$111,482
Gemini 3.6 FlashGoogle$31,482
Claude Opus 4.8Anthropic$101,481
MiMo-V2.6-ProXiaomi MiMo$0.541,480
GLM-5.3Z.ai$2.151,478
Claude Sonnet 5.5Anthropic$41,476
GPT-5.4OpenAI$5.631,475
GPT-6 AstraOpenAI$201,475
DeepSeek-V4.1-FlashDeepSeek$0.521,475
GLM-5.3-FlashZ.ai$0.241,475
Gemini 3 Flash PreviewGoogle$1.131,473
Grok 4.20xAI$1.561,472
GPT-5.6 TerraOpenAI$4.501,466
Grok 4.5xAI$31,466
DeepSeek-V4-ProDeepSeek$1.981,465
Claude Sonnet 5Anthropic$41,461
GPT-6 SolOpenAI$41,456
Gemini 3.5 Flash-LiteGoogle$0.851,455
Grok 4.6xAI$31,454
GPT-5.6 LunaOpenAI$0.451,452
MiMo-V2.6-FlashXiaomi MiMo$0.181,451
GPT-6 LunaOpenAI$0.201,443
Grok 4.7xAI$31,443
Grok 4.3xAI$1.561,442
Mistral Large 4Mistral AI$2.061,429
GPT-4.1OpenAI$3.501,415
Solar Pro 4Upstage$0.521,386
GPT-4.1 miniOpenAI$0.701,383
GPT-4.1 nanoOpenAI$0.181,322

Cost

What each costs

Example workloads at each provider's listed rates, then every other rate either one lists. The cheaper figure in each row is highlighted.

The same work, priced

For 10M input and 2M output tokens, Grok 4.6 costs 7.3 times as much as GPT-5.6 Luna, at each provider's listed rates.

What GPT-5.6 Luna and Grok 4.6cost for the same work, at each provider's listed rates.

  • 10M input and 2M output tokens

    Grok 4.6 costs 7.3× as much

    GPT-5.6 Luna
    $4.40
    Grok 4.6
    $32

  • The same, with 9M of the input tokens cached

    Grok 4.6 costs 6.7× as much

    GPT-5.6 Luna
    $2.78
    Grok 4.6
    $18.50
Workload or rateGPT-5.6 LunaGrok 4.6
Example workloads
10M input and 2M output tokens$4.40 (better)$32
The same, with 9M of the input tokens cached$2.78 (better)$18.50
Other listed rates
1M input tokens, cached$0.02 (better)$0.50

Verdict

When to choose each

GPT-5.6 Luna

Where it leads

  • Lower input and output prices, $0.20 and $1.20 vs $2 and $6 per 1M tokens
  • Higher FrontierMath Tiers 1-3 (v2) score, 82.1% vs 66%
  • Larger context window 1,050,000 vs 500,000 tokens
  • Cheaper cached input, $0.02 vs $0.50 per 1M tokens

Best for

  • High-volume API evaluation with published low token prices
  • Subagents and batch reasoning where cost per call dominates

Watch out for

  • Not in the pinned LiveBench snapshot; its independent scores here come from Epoch AI's own GPQA Diamond, FrontierMath and SimpleQA Verified runs
Full GPT-5.6 Luna profile

Grok 4.6

Where it leads

  • Higher Arena Text score, 1454 Elo vs 1452 Elo
  • Higher GPQA Diamond score, 94% vs 91.6%
  • Higher SimpleQA Verified score, 49.3% vs 41%

Best for

  • Coding, agentic, and knowledge-work API evaluations with controlled long-context cost

Watch out for

  • Published token rates increase for all tokens in a request once its prompt reaches xAI's 200,000-token long-context threshold
Full Grok 4.6 profile

From benchmark to production

Building on GPT-5.6 Luna or Grok 4.6?

Public scores rarely hold on your own data. We test both models on your real tasks, then build the agent or pipeline around the one that wins.

Book a consultationAI agent development

FAQ

GPT-5.6 Luna vs Grok 4.6 FAQ

Which is cheaper, GPT-5.6 Luna or Grok 4.6?

GPT-5.6 Luna is 90% cheaper per 1M input tokens than Grok 4.6, $0.20 vs $2. For 10M input and 2M output tokens, GPT-5.6 Luna costs $4.40 and Grok 4.6 $32 at standard rates.

Which is cheaper with prompt caching, GPT-5.6 Luna or Grok 4.6?

GPT-5.6 Luna charges $0.02 per 1M cached input tokens, 90% below its standard input rate, and Grok 4.6 $0.50, 75% below. With 9M of 10M input tokens cached and 2M output tokens, GPT-5.6 Luna costs $2.78 and Grok 4.6 $18.50.

Which scores higher on independent benchmarks, GPT-5.6 Luna or Grok 4.6?

Grok 4.6 leads on Arena Text, 1454 Elo to 1452 Elo; Grok 4.6 leads on GPQA Diamond, 94% to 91.6%; GPT-5.6 Luna leads on FrontierMath Tiers 1-3 (v2), 82.1% to 66%. Scores are independent evaluations, not either vendor's launch figures.

Which has the larger context window, GPT-5.6 Luna or Grok 4.6?

GPT-5.6 Luna has the larger context window, 1,050,000 vs 500,000 tokens.

Further reading

More on GPT-5.6 Luna and Grok 4.6

Articles on GPT-5.6 Luna and Grok 4.6

Related comparisons

Production evaluation

Choosing between GPT-5.6 Luna and Grok 4.6?

Public benchmarks narrow the shortlist. We run both on your own tasks, data and cost limits, then build the routing and review controls to run the winner in production.

  • A head-to-head evaluation on your real workload
  • Routing between models by cost and difficulty
Compare all modelsBook a consultation