Model comparison
GPT-5.6 Luna vs MAI-Thinking-1
Head-to-head evidence from 5 shared benchmark results across 3 categories. Overall scores shown here use the public BenchAlign v5 ranking lane.
Public leaderboard positions: GPT-5.6 Luna #22 (Estimated); MAI-Thinking-1 unranked (Not scored). Intervals and evidence labels describe ranking uncertainty, not a guarantee for a specific workload.
Evidence parity. GPT-5.6 Luna and MAI-Thinking-1 share 5 comparable benchmark results. 4 of 8 categories are comparable. 36 results are unique to GPT-5.6 Luna; 8 to MAI-Thinking-1.
Updated July 23, 2026- Shared results
- 5
- GPT-5.6 Luna only
- 36
- MAI-Thinking-1 only
- 8
- Comparable categories
- 4 / 8
Treat this as a split decision. GPT-5.6 Luna makes more sense if agentic is the priority or you need the larger 1M context window; MAI-Thinking-1 is the better fit if mathematics is the priority.
Confidence note. This is a partial-evidence comparison with 5 shared benchmark results across 3 evidence categories; 4 of 8 categories currently have scoreable aggregates for both models. Treat the verdict as directional until coverage is more balanced.
Why this result
GPT-5.6 Luna and MAI-Thinking-1 finish on the same BenchAlign overall score, so this is less about a single winner and more about where the edge shows up. The BenchAlign headline says tie; the benchmark table is where the real choice happens.
GPT-5.6 Luna gives you the larger context window at 1M, compared with 256K for MAI-Thinking-1.
Category breakdown
Exact category averages are shown below. Not measured means BenchLM does not have enough sourced public coverage for that model and category.
| Category | GPT-5.6 Luna | Δ | MAI-Thinking-1 |
|---|---|---|---|
| Agentic | GPT-5.6 Luna84.1 | Margin← 38.1 | MAI-Thinking-146.0 |
| Knowledge | GPT-5.6 Luna92.3 | Margin← 19.8 | MAI-Thinking-172.5 |
| Math | GPT-5.6 Luna73.6 | Margin→ 16.1 | MAI-Thinking-189.7 |
| Coding | GPT-5.6 Luna62.7 | Margin→ 2.8 | MAI-Thinking-165.5 |
| Multimodal | GPT-5.6 Luna78.4 | MarginNo overlap | MAI-Thinking-1Not measured |
| Inst. Following | GPT-5.6 LunaNot measured | MarginNo overlap | MAI-Thinking-185.0 |
Decisive benchmark drivers
The largest measured benchmark gaps in this matchup, with exact reported values.
More
- Source ↗
Terminal-Bench 2.0
AgenticA 84.7%B 46%Winner: GPT-5.6 LunaΔ 38.7Terminal-Bench 2.0: GPT-5.6 Luna scored 84.7%; MAI-Thinking-1 scored 46%. GPT-5.6 Luna wins this benchmark. - Source ↗
SWE-bench Pro
CodingA 62.7%B 52.8%Winner: GPT-5.6 LunaΔ 9.9SWE-bench Pro: GPT-5.6 Luna scored 62.7%; MAI-Thinking-1 scored 52.8%. GPT-5.6 Luna wins this benchmark. - Source ↗
GPQA
KnowledgeA 92.3%B 84.2%Winner: GPT-5.6 LunaΔ 8.1GPQA: GPT-5.6 Luna scored 92.3%; MAI-Thinking-1 scored 84.2%. GPT-5.6 Luna wins this benchmark.
Operational comparison
Runtime and commercial metrics are compared only when both models have a complete sourced value.
| Metric | GPT-5.6 Luna | MAI-Thinking-1 | Comparison |
|---|---|---|---|
| Input / output priceUSD per 1M tokens | GPT-5.6 Luna$1 input / $6 output | MAI-Thinking-1Not available | A complete price comparison is not available. |
| Generation speedtokens per second | GPT-5.6 LunaNot available | MAI-Thinking-1Not available | A complete speed comparison is not available. |
| First-answer latencyseconds to first token | GPT-5.6 LunaNot available | MAI-Thinking-1Not available | A complete latency comparison is not available. |
| Context windowmaximum listed tokens | GPT-5.6 Luna1M | MAI-Thinking-1256K | GPT-5.6 Luna lists the larger context window. |
Benchmark Deep Dive
AgenticGPT-5.6 Luna wins15 benchmarks
| Benchmark | GPT-5.6 Luna | MAI-Thinking-1 | Result |
|---|---|---|---|
| Terminal-Bench 2.0Source | 84.7% | 46% | GPT-5.6 Luna leads |
| BrowseCompSource | 83.3% | — | Not comparable |
| OSWorld 2.0Source | 45.6% | — | Not comparable |
| CyberGymSource | 77.9% | — | Not comparable |
| ExploitGymSource | 12.4% | — | Not comparable |
| ToolathlonSource | 53.4% | — | Not comparable |
| AA Agentic IndexSource | 45.6% | — | Not comparable |
| GDPval-AASource | 54.2% | — | Not comparable |
| GDPval-AASource | 1584 | — | Not comparable |
| AA Harvey LABSource | 87.9% | — | Not comparable |
| AA ITBenchSource | 40.3% | — | Not comparable |
| AA Tau3 BankingSource | 27.2% | — | Not comparable |
| AA AutomationBenchSource | 42.2% | — | Not comparable |
| aaTerminalBench21Source | 80.9% | — | Not comparable |
| APEX-Agents-AASource | 35.8% | — | Not comparable |
CodingMAI-Thinking-1 wins8 benchmarks
| Benchmark | GPT-5.6 Luna | MAI-Thinking-1 | Result |
|---|---|---|---|
| SWE-bench ProSource | 62.7% | 52.8% | GPT-5.6 Luna leads |
| Terminal-Bench 2.0Source | 84.7% | 46.0% | GPT-5.6 Luna leads |
| deepSweSource | 67.2% | — | Not comparable |
| FrontierCode 1.1 ExtendedSource | 55.1% | — | Not comparable |
| cursorBench32Source | 61.1% | — | Not comparable |
| AA Coding IndexSource | 71.5% | — | Not comparable |
| AA-SciCodeSource | 52.5% | — | Not comparable |
| SWE-bench VerifiedSource | — | 73.5% | Not comparable |
Reasoning4 benchmarks
KnowledgeGPT-5.6 Luna wins12 benchmarks
| Benchmark | GPT-5.6 Luna | MAI-Thinking-1 | Result |
|---|---|---|---|
| GPQASource | 92.3% | 84.2% | GPT-5.6 Luna leads |
| GPQA-DSource | 92.3% | 84.2% | GPT-5.6 Luna leads |
| HealthBench ProfessionalSource | 55.7% | — | Not comparable |
| HealthBench HardSource | 32.0% | — | Not comparable |
| Artificial Analysis Intelligence IndexSource | 51.2% | — | Not comparable |
| AA-GPQA DiamondSource | 91.1% | — | Not comparable |
| AA-HLESource | 37.2% | — | Not comparable |
| AA-Omniscience IndexSource | -11.2% | — | Not comparable |
| AA-Omniscience AccuracySource | 41.5% | — | Not comparable |
| AA-Omniscience Hallucination RateSource | 90.1% | — | Not comparable |
| MMLU-ProSource | — | 85% | Not comparable |
| SimpleQASource | — | 31% | Not comparable |
MathMAI-Thinking-1 wins6 benchmarks
| Benchmark | GPT-5.6 Luna | MAI-Thinking-1 | Result |
|---|---|---|---|
| FrontierMath (legacy)Source | 78.6% | — | Not comparable |
| FrontierMath v2 (Tiers 1-3)Source | 78.600% | — | Not comparable |
| FrontierMath v2 (Tier 4)Source | 58.500% | — | Not comparable |
| AIME 2025Source | — | 97% | Not comparable |
| AIME26Source | — | 94.5% | Not comparable |
| HMMT Feb 2026Source | — | 84.9% | Not comparable |
Multimodal3 benchmarks
Inst. Following1 benchmarks
| Benchmark | GPT-5.6 Luna | MAI-Thinking-1 | Result |
|---|---|---|---|
| IFBenchSource | — | 85% | Not comparable |
Frequently Asked Questions (5)
Which is better, GPT-5.6 Luna or MAI-Thinking-1?
GPT-5.6 Luna and MAI-Thinking-1 are tied on the BenchAlign overall score, so the right pick depends on which category matters most for your use case.
Which is better for knowledge tasks, GPT-5.6 Luna or MAI-Thinking-1?
GPT-5.6 Luna has the edge for knowledge tasks in this comparison, averaging 92.3 versus 72.5. Inside this category, GPQA is the benchmark that creates the most daylight between them.
Which is better for coding, GPT-5.6 Luna or MAI-Thinking-1?
MAI-Thinking-1 has the edge for coding in this comparison, averaging 65.5 versus 62.7. Inside this category, Terminal-Bench 2.0 is the benchmark that creates the most daylight between them.
Which is better for math, GPT-5.6 Luna or MAI-Thinking-1?
MAI-Thinking-1 has the edge for math in this comparison, averaging 89.7 versus 73.6. GPT-5.6 Luna stays close enough that the answer can still flip depending on your workload.
Which is better for agentic tasks, GPT-5.6 Luna or MAI-Thinking-1?
GPT-5.6 Luna has the edge for agentic tasks in this comparison, averaging 84.1 versus 46. Inside this category, Terminal-Bench 2.0 is the benchmark that creates the most daylight between them.
Related Comparisons
Explore More
Choose a model with this week’s evidence
Join 2,000+ readers for ranking moves, pricing changes, and the claims that still need proof.
One email each week. Unsubscribe anytime.