GPT-6 Luna
| Released | 2026-09-22 |
|---|---|
| Price per 1M | $0.10 in · $0.50 out |
| Overall score | 58.0 · rank 53 |
| Context window | 1,050,000 tokens |
| Provider | OpenAI |
| Weights | Proprietary |
Scores
| Dimension | Score | Rank | Benchmarks |
|---|---|---|---|
| Overall: Balanced | 58.0 | 53 | 45 |
| Overall: Coding | 59.1 | 38 | 37 |
| Overall: Reasoning | 58.7 | 49 | 18 |
| Overall: Agentic work | 58.9 | 44 | 34 |
| Overall: Frontend | 58.2 | 37 | 13 |
| Coding | 59.5 | 31 | 9 |
| Agents | 58.6 | 43 | 13 |
| Reasoning | 58.9 | 56 | 12 |
| Math | 61.2 | 22 | 4 |
| Knowledge | 54.1 | 69 | 2 |
| Frontend | 58.3 | 37 | 3 |
| General | 54.3 | 72 | 1 |
| Multimodal | 53.4 | 59 | 1 |
| Value vs price | +15.0 | 7 |
Configurations
| Configuration | Balanced | Benchmarks |
|---|---|---|
| Unspecified setting | 59.0 ±2.8 | 8 |
| High | 56.0 ±4.8 | 6 |
| Low | 51.6 ±4.9 | 5 |
| Max | 58.0 ±0.7 | 36 |
| Medium | 55.7 ±4.9 | 5 |
| Non-reasoning | 47.1 ±4.9 | 5 |
| Extra High | 57.9 ±4.2 | 6 |
Evidence
Coding 59.5 ±1.0
| Benchmark | Result | Source | Date |
|---|---|---|---|
| FrontierCode | 42.4% | Epoch AI | 2026-11-11 |
| Code Migration v1.1 | 46.7% | Vals AI | 2026-11-09 |
| LiveBench 2.0 Coding | 70.8% | LiveBench | 2026-11-06 |
| WeirdML v3 | 7.9% | Epoch AI | 2026-11-11 |
| MirrorCode | 26.7% | Epoch AI | 2026-10-21 |
| IOI | 55.6% | Vals AI | 2026-10-29 |
| SWE-bench Multimodal | 37.5% | Vals AI | 2026-11-09 |
| SciCode | 54.6% | Epoch AI | 2026-11-11 |
| ProgramBench | 0.5% | Vals AI | 2026-11-09 |
Agents 58.6 ±1.1
| Benchmark | Result | Source | Date |
|---|---|---|---|
| Terminal-Bench 4 | 15.2% | Terminal-Bench | 2026-09-22 |
| SRE Bench | 2.7% | Vals AI | 2026-11-05 |
| Terminal-Bench-Science | 4.3% | Vals AI | 2026-11-09 |
| Terminal-Bench 4.0 (Vals run) | 13.6% | Vals AI | 2026-11-09 |
| APEX-Agents | 44.3% | Epoch AI | 2026-11-11 |
| LiveBench 2.0 Agentic | 24.7% | LiveBench | 2026-11-06 |
| LiveBench 2.0 Browser Use | 70.3% | LiveBench | 2026-11-06 |
| Terminal-Bench 2.1 (Vals run) | 73.0% | Vals AI | 2026-10-01 |
| EMB v1.1 (Excel financial modeling) | 70.0% | Vals AI | 2026-11-09 |
| LiveBench 2.0 Research | 48.5% | LiveBench | 2026-11-06 |
| Tax Agent Bench | 19.9% | Vals AI | 2026-11-09 |
| Finance Agent v2.1 | 56.8% | Vals AI | 2026-11-09 |
| HLAB-Vals 1.1 (legal agent) | 4.2% | Vals AI | 2026-11-09 |
| CyberBench v1.1 | 76.3% | Vals AI | 2026-11-05 |
Reasoning 58.9 ±1.6
| Benchmark | Result | Source | Date |
|---|---|---|---|
| ARC-AGI-2 | 59.3% | Epoch AI | 2026-11-11 |
| ARC-AGI-2 (ARC Prize verified) | 59.3% | ARC Prize | 2026-11-11 |
| Extended NYT Connections | 68.7% | Lech Mazur | 2026-11-10 |
| MysteryMechanism | 19.4% | Vals AI | 2026-11-09 |
| GPQA Diamond | 90.5% | Epoch AI | 2026-09-22 |
| CritPt | 19.4% | Epoch AI | 2026-11-11 |
| ARC-AGI-1 (ARC Prize verified) | 86.7% | ARC Prize | 2026-11-11 |
| ARC-AGI-1 | 86.7% | Epoch AI | 2026-11-11 |
| BioMysteryBench | 61.5% | Vals AI | 2026-11-09 |
| ARC-AGI-3 | 0.6% | ARC Prize | 2026-11-11 |
| Chess Puzzles | 36.0% | Epoch AI | 2026-10-13 |
| LiveBench 2.0 Reasoning | 81.8% | LiveBench | 2026-11-06 |
| Mystery Game Puzzles | 7.0% | Epoch AI | 2026-09-22 |
| LiveBench 2.0 Data Analysis | 73.4% | LiveBench | 2026-11-06 |
Math 61.2 ±1.2
| Benchmark | Result | Source | Date |
|---|---|---|---|
| FrontierMath Tiers 1-3 | 79.0% | Epoch AI | 2026-09-22 |
| FrontierMath Tier 4 | 56.1% | Epoch AI | 2026-09-22 |
| OTIS Mock AIME 2024-2025 | 98.9% | Epoch AI | 2026-09-22 |
| ProofBench v1.1 | 64.0% | Vals AI | 2026-11-09 |
| ProofBench | 64.0% | Epoch AI | 2026-11-11 |
Knowledge 54.1 ±2.7
| Benchmark | Result | Source | Date |
|---|---|---|---|
| SimpleQA Verified | 41.4% | Epoch AI | 2026-09-22 |
| Legal Research Bench | 30.3% | Vals AI | 2026-11-09 |
Frontend 58.3 ±2.2
| Benchmark | Result | Source | Date |
|---|---|---|---|
| LiveBench 2.0 Design | 76.1% | LiveBench | 2026-11-06 |
| Vibe Code Bench v1.1 | 81.6% | Vals AI | 2026-11-09 |
| LMArena WebDev | 1,578 | LMArena | 2026-11-09 |
General 54.3 ±4.1
| Benchmark | Result | Source | Date |
|---|---|---|---|
| LMArena Text | 1,441 | LMArena | 2026-11-09 |
Multimodal 53.4 ±4.1
| Benchmark | Result | Source | Date |
|---|---|---|---|
| LMArena Vision | 1,235 | LMArena | 2026-11-09 |
Results not used in the score
| Benchmark | Result | Source | Why not used |
|---|---|---|---|
| ARC-AGI-1 | 73.0% | Epoch AI | A stronger configuration result was used |
| ARC-AGI-1 | 61.0% | Epoch AI | A stronger configuration result was used |
| ARC-AGI-1 | 70.3% | Epoch AI | A stronger configuration result was used |
| ARC-AGI-1 | 37.7% | Epoch AI | A stronger configuration result was used |
| ARC-AGI-1 | 8.8% | Epoch AI | A stronger configuration result was used |
| ARC-AGI-1 (ARC Prize verified) | 73.0% | ARC Prize | A stronger configuration result was used |
| ARC-AGI-1 (ARC Prize verified) | 61.0% | ARC Prize | A stronger configuration result was used |
| ARC-AGI-1 (ARC Prize verified) | 70.3% | ARC Prize | A stronger configuration result was used |
| ARC-AGI-1 (ARC Prize verified) | 37.7% | ARC Prize | A stronger configuration result was used |
| ARC-AGI-1 (ARC Prize verified) | 8.8% | ARC Prize | A stronger configuration result was used |
| ARC-AGI-2 | 41.9% | Epoch AI | A stronger configuration result was used |
| ARC-AGI-2 | 18.1% | Epoch AI | A stronger configuration result was used |
| ARC-AGI-2 | 31.4% | Epoch AI | A stronger configuration result was used |
| ARC-AGI-2 | 4.6% | Epoch AI | A stronger configuration result was used |
| ARC-AGI-2 | 0.0% | Epoch AI | A stronger configuration result was used |
| ARC-AGI-2 (ARC Prize verified) | 41.9% | ARC Prize | A stronger configuration result was used |
| ARC-AGI-2 (ARC Prize verified) | 18.1% | ARC Prize | A stronger configuration result was used |
| ARC-AGI-2 (ARC Prize verified) | 31.4% | ARC Prize | A stronger configuration result was used |
| ARC-AGI-2 (ARC Prize verified) | 4.6% | ARC Prize | A stronger configuration result was used |
| ARC-AGI-2 (ARC Prize verified) | 0.0% | ARC Prize | A stronger configuration result was used |
| ARC-AGI-3 | 0.1% | ARC Prize | A stronger configuration result was used |
| ARC-AGI-3 | 0.2% | ARC Prize | A stronger configuration result was used |
| ARC-AGI-3 | 0.5% | ARC Prize | A stronger configuration result was used |
| ARC-AGI-3 | 0.2% | ARC Prize | A stronger configuration result was used |
| ARC-AGI-3 | 0.3% | ARC Prize | A stronger configuration result was used |
| ARC-AGI-3 | 0.2% | ARC Prize | A stronger configuration result was used |
| ARC-AGI-3 | 0.4% | ARC Prize | A stronger configuration result was used |
| ARC-AGI-3 | 0.0% | ARC Prize | A stronger configuration result was used |
| ARC-AGI-3 | 0.2% | ARC Prize | A stronger configuration result was used |
| ARC-AGI-3 | 0.0% | ARC Prize | A stronger configuration result was used |
| ARC-AGI-3 | 0.1% | ARC Prize | A stronger configuration result was used |
| CritPt | 17.4% | Epoch AI | A stronger configuration result was used |
| CritPt | 10.6% | Epoch AI | A stronger configuration result was used |
| CritPt | 15.4% | Epoch AI | A stronger configuration result was used |
| CritPt | 2.6% | Epoch AI | A stronger configuration result was used |
| CritPt | 1.1% | Epoch AI | A stronger configuration result was used |
| SciCode | 51.7% | Epoch AI | A stronger configuration result was used |
| SciCode | 50.9% | Epoch AI | A stronger configuration result was used |
| SciCode | 50.4% | Epoch AI | A stronger configuration result was used |
| SciCode | 46.9% | Epoch AI | A stronger configuration result was used |
| SciCode | 43.1% | Epoch AI | A stronger configuration result was used |
Names matched to this model
gpt-6-luna