# GPT-6 Astra vs GPT-6 Luna

> Side-by-side benchmark comparison for GPT-6 Astra and GPT-6 Luna across agentic, coding, multimodal, knowledge, reasoning, multilingual, and math tasks.

- Canonical page: https://benchlm.ai/compare/gpt-6-astra-vs-gpt-6-luna
- Last updated: September 22, 2026

- Shared sourced benchmarks: 7
- HTML indexing: indexable
- Ranking lane: BenchAlign v5

## Quick Verdict

Pick GPT-6 Astra if you want the stronger benchmark profile. GPT-6 Luna only makes more sense when its price, context window, or workload-specific category wins matter more than the overall score.

## Summary

- GPT-6 Astra leads overall 82.93 to 62.33.
- The clearest category separation is in knowledge, where the averages are 81.8 for GPT-6 Astra and 63.6 for GPT-6 Luna.
- The biggest single benchmark swing is ExploitGym in Agentic, with scores of 42.4% and 11.6%.
- GPT-6 Luna is the cheaper option on output tokens, which matters if you expect large responses or heavy interactive use.

## Model Snapshot

| Property | GPT-6 Astra | GPT-6 Luna |
|----------|----------|----------|
| Creator | OpenAI | OpenAI |
| Type | Proprietary | Proprietary |
| Reasoning | Reasoning | Reasoning |
| Context | 1.05M | 1.05M |
| Overall Score | 82.93 | 62.33 |
| Benchmarks Covered | 30 | 7 |
| Pricing (input/output) | $10.00 / $50.00 | $0.10 / $0.50 |

## Category Breakdown

### Agentic

- Winner: Coming soon
- GPT-6 Astra public-lane score: 70.5 (Supported · #6/157)
- GPT-6 Luna public-lane score: 56.7 (Estimated · #35/157)

| Benchmark | GPT-6 Astra | GPT-6 Luna | Winner |
|-----------|-----------|-----------|--------|
| BrowseComp | 91.5% | Coming soon | Coming soon |
| OSWorld 2.0 | 72.6% | Coming soon | Coming soon |
| Terminal-Bench 4.0 | 57.90% | Coming soon | Coming soon |
| Terminal-Bench-Science 0.1 | 64.6% | Coming soon | Coming soon |
| ExploitGym | 42.4% | 11.6% | GPT-6 Astra |
| Agents' Last Exam | 59.3% | Coming soon | Coming soon |
| AutomationBench | 41.4% | Coming soon | Coming soon |
| HLE w/ tools | 57.2% | Coming soon | Coming soon |
| Terminal-Bench 2.1 (Vals) | 87.3% | Coming soon | Coming soon |
| ApprenticeBench | 68% | Coming soon | Coming soon |

### Coding

- Winner: Coming soon
- GPT-6 Astra public-lane score: 76.3 (Supported · #4/159)
- GPT-6 Luna public-lane score: 66.6 (Estimated · #11/159)

| Benchmark | GPT-6 Astra | GPT-6 Luna | Winner |
|-----------|-----------|-----------|--------|
| DeepSWE | 74.1% | 66.6% | GPT-6 Astra |
| FrontierCode 1.1 Main | 53.3% | Coming soon | Coming soon |
| FrontierCode 1.1 Extended | 64.5% | Coming soon | Coming soon |
| FrontierSWE v2 | 65.5% | Coming soon | Coming soon |

### Reasoning

- Winner: Coming soon
- GPT-6 Astra public-lane score: 89.5 (#1/17)
- GPT-6 Luna public-lane score: 78.3 (Unranked · 2 rankable rows)

| Benchmark | GPT-6 Astra | GPT-6 Luna | Winner |
|-----------|-----------|-----------|--------|
| ARC-AGI-1 | 98.50% | Coming soon | Coming soon |
| ARC-AGI-2 | 95% | Coming soon | Coming soon |
| ARC-AGI-3 | 62.7% | Coming soon | Coming soon |
| GeneBench-Pro | 37.8% | Coming soon | Coming soon |
| MRCR v2 256K-512K | 100.0% | Coming soon | Coming soon |
| MRCR v2 512K-1M | 96.3% | Coming soon | Coming soon |

### Multimodal & Grounded

- Winner: Coming soon
- GPT-6 Astra public-lane score: 82.9 (Unranked · 3 rankable rows)
- GPT-6 Luna public-lane score: 71.3 (Unranked · 1 rankable row)

| Benchmark | GPT-6 Astra | GPT-6 Luna | Winner |
|-----------|-----------|-----------|--------|
| ScreenSpot Pro | 92.7% | Coming soon | Coming soon |
| BenchCAD Vision2Code (tools) | 0.959 | Coming soon | Coming soon |

### Knowledge

- Winner: Coming soon
- GPT-6 Astra public-lane score: 81.8 (Supported · #3/189)
- GPT-6 Luna public-lane score: 63.6 (Estimated · #29/189)

| Benchmark | GPT-6 Astra | GPT-6 Luna | Winner |
|-----------|-----------|-----------|--------|
| GPQA | 96% | Coming soon | Coming soon |
| GPQA-D | 96.0% | Coming soon | Coming soon |
| HealthBench (raw) | 56.9% | 50.0% | GPT-6 Astra |
| HealthBench (length-adjusted) | 58.3% | 54.5% | GPT-6 Astra |
| HealthBench Professional | 64.7% | 60.8% | GPT-6 Astra |
| HealthBench Professional (raw) | 68.2% | 61.2% | GPT-6 Astra |
| HealthBench Hard | 36.6% | 31.4% | GPT-6 Astra |

### Mathematics

- Winner: Coming soon
- GPT-6 Astra public-lane score: 85 (Unranked · 1 rankable row)
- GPT-6 Luna public-lane score: Coming soon

| Benchmark | GPT-6 Astra | GPT-6 Luna | Winner |
|-----------|-----------|-----------|--------|
| FrontierMath v2 (Tier 4) | 97.600% | Coming soon | Coming soon |

## FAQ

### Which is better overall, GPT-6 Astra or GPT-6 Luna?

GPT-6 Astra is ahead overall on BenchLM right now.

### Where is the biggest gap between GPT-6 Astra and GPT-6 Luna?

The widest category gap is in knowledge, where the averages are 81.8 for GPT-6 Astra and 63.6 for GPT-6 Luna.

### How many benchmarks does GPT-6 Astra cover on BenchLM?

GPT-6 Astra currently has 30 sourced benchmark scores on BenchLM.

### How many benchmarks does GPT-6 Luna cover on BenchLM?

GPT-6 Luna currently has 7 sourced benchmark scores on BenchLM.

## Related Comparisons

- [GPT-6 Astra vs GPT-6 Sol](/compare/gpt-6-astra-vs-gpt-6-sol)
- [GPT-6 Luna vs GPT-6 Sol](/compare/gpt-6-luna-vs-gpt-6-sol)
- [GPT-6 Astra vs Claude Fable 5.1](/compare/claude-fable-5-1-vs-gpt-6-astra)
- [GPT-6 Luna vs Claude Fable 5.1](/compare/claude-fable-5-1-vs-gpt-6-luna)

## Explore More

- [GPT-6 Astra profile](/models/gpt-6-astra)
- [GPT-6 Luna profile](/models/gpt-6-luna)
- [Compare Pricing](/llm-pricing)
- [Alternative Finder](/tools/alternative-finder)
- [LLM Selector](/tools/llm-selector)
- [Overall Rankings](/best/overall)