Skip to content
VECTOR WIREAI INTELLIGENCE
UTC

GPT-6 Astra vs GPT-6 Luna

25 SHARED BENCHMARKS

Across 25 shared benchmarks, GPT-6 Astra scores higher on 24 and GPT-6 Luna on 1. The widest gap is Terminal-Bench 4.0, where GPT-6 Astra scores 59.1 against 12.6. GPT-6 Luna is the cheaper of the two on tracked API pricing ($0.10 against $10.00 per million input tokens).

OPENAIVSOPENAI25 SHARED24–1 HEAD-TO-HEADNEWEST SCORE ADDED

At a glance

MakerOpenAIOpenAI
Released
Price per 1M tokens input / output$10.00 / $50.00$0.10 / $0.50
Cost of 1M in + 1M out$60.00$0.60 100.0× less
Head-to-head of 25 shared benchmarks24 wins1 wins
Scores tracked independently verified160 43 ◆102 33 ◆

Release dates per Artificial Analysis. Prices: Artificial Analysis for GPT-6 Astra; Artificial Analysis for GPT-6 Luna. ◆ = independently verified score.

Where each leadsby capability area · benchmark wins

AreaGPT-6 AstraWINSGPT-6 Luna
Reasoning50GPT-6 Astra leads 5 of 5 · widest: CritPt 31.7 vs 19.4
Coding40GPT-6 Astra leads 4 of 4 · widest: LMArena · WebDev 1792 vs 1593
Agentic20GPT-6 Astra leads 2 of 2 · widest: Terminal-Bench 4.0 59.1 vs 12.6
Factuality20GPT-6 Astra leads 2 of 2 · widest: AA-Omniscience · Non-hallucination 48.7 vs 23.3
Instruction Following10GPT-6 Astra leads 1 of 1 · widest: LiveBench · Instruction Following 75.6 vs 55.9
Long Context01GPT-6 Luna leads 1 of 1
Math10GPT-6 Astra leads 1 of 1 · widest: LiveBench · Mathematics 96.8 vs 89.1
Multimodal10GPT-6 Astra leads 1 of 1 · widest: MMMU-Pro 86.9 vs 75.5

Biggest gaps

GPT-6 Astra pulls furthest ahead on

  1. Terminal-Bench 4.059.1 vs 12.6
  2. AA-Omniscience · Non-hallucination48.7 vs 23.3
  3. CritPt31.7 vs 19.4

GPT-6 Luna pulls furthest ahead on

No ratified-area lead of 3 points or more.

Every shared benchmark25 · grouped by area

Reasoning 5

BenchmarkGPT-6 AstraMARGINGPT-6 Luna
ARC-AGI-399.9◆ 0.1
ARC-AGI-295◆ 59.3
CritPt31.719.4
LiveBench · Reasoning92.7 ◆◆ 81.8

Coding 4

BenchmarkGPT-6 AstraMARGINGPT-6 Luna
LMArena · WebDev1792 ◆◆ 1593
LiveBench · Coding80.4 ◆◆ 79
SciCode56.554.6

Agentic 2

BenchmarkGPT-6 AstraMARGINGPT-6 Luna
GDPVal52.143.4
Terminal-Bench 4.059.112.6

Instruction Following 1

Long Context 1

BenchmarkGPT-6 AstraMARGINGPT-6 Luna
AA-LCR80.783.3

Math 1

BenchmarkGPT-6 AstraMARGINGPT-6 Luna
LiveBench · Mathematics96.8 ◆◆ 89.1

Multimodal 1

BenchmarkGPT-6 AstraMARGINGPT-6 Luna
MMMU-Pro86.975.5

Other shared benchmarks 8

BenchmarkGPT-6 AstraMARGINGPT-6 Luna
ARC-AGI-197.5 ◆◆ 86.7
DeepSWE 1.174.166.6
livebench_language89.4 ◆◆ 73.8
OSWorld 2.072.652.7

Each score is the one the model's own page shows — the most authoritative tracked result for that benchmark, on the benchmark's standard methodology (source-attributed). ↓ marks lower-is-better metrics; ◆ an independently verified score. Open either model for its full surface, provenance and pricing. Quoted rates are the price-setter row we currently track for each model — its direct or vendor-official listing where one exists (direct, direct), otherwise the lowest tracked offer.

Compare GPT-6 Astra withALL PAIRINGS →