Model comparison
Hy3 Preview vs MAI-Thinking-1
Head-to-head evidence from 6 shared benchmark results across 4 categories. Overall scores shown here use the public BenchAlign v5 ranking lane.
Evidence parity. Hy3 Preview and MAI-Thinking-1 share 6 comparable benchmark results. 4 of 8 categories are comparable. 16 results are unique to Hy3 Preview; 7 to MAI-Thinking-1.
Updated July 23, 2026- Shared results
- 6
- Hy3 Preview only
- 16
- MAI-Thinking-1 only
- 7
- Comparable categories
- 4 / 8
Treat this as a split decision. Hy3 Preview makes more sense if agentic is the priority; MAI-Thinking-1 is the better fit if knowledge is the priority.
Confidence note. This is a partial-evidence comparison with 6 shared benchmark results across 4 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
Hy3 Preview 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.
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 | Hy3 Preview | Δ | MAI-Thinking-1 |
|---|---|---|---|
| Knowledge | Hy3 Preview33.8 | Margin→ 38.7 | MAI-Thinking-172.5 |
| Inst. Following | Hy3 Preview63.1 | Margin→ 21.9 | MAI-Thinking-185.0 |
| Agentic | Hy3 Preview54.4 | Margin← 8.4 | MAI-Thinking-146.0 |
| Coding | Hy3 Preview57.8 | Margin→ 7.7 | MAI-Thinking-165.5 |
| Math | Hy3 PreviewNot measured | MarginNo overlap | MAI-Thinking-189.7 |
Decisive benchmark drivers
The largest measured benchmark gaps in this matchup, with exact reported values.
More
- Source ↗
IFBench
Inst. FollowingA 63.1%B 85%Winner: MAI-Thinking-1Δ 21.9IFBench: Hy3 Preview scored 63.1%; MAI-Thinking-1 scored 85%. MAI-Thinking-1 wins this benchmark. - Source ↗
Terminal-Bench 2.0
AgenticA 54.4%B 46%Winner: Hy3 PreviewΔ 8.4Terminal-Bench 2.0: Hy3 Preview scored 54.4%; MAI-Thinking-1 scored 46%. Hy3 Preview wins this benchmark. - Source ↗
GPQA
KnowledgeA 87.2%B 84.2%Winner: Hy3 PreviewΔ 3GPQA: Hy3 Preview scored 87.2%; MAI-Thinking-1 scored 84.2%. Hy3 Preview wins this benchmark. - Source ↗
SWE-bench Verified
CodingA 74.4%B 73.5%Winner: Hy3 PreviewΔ 0.9SWE-bench Verified: Hy3 Preview scored 74.4%; MAI-Thinking-1 scored 73.5%. Hy3 Preview wins this benchmark.
Operational comparison
Runtime and commercial metrics are compared only when both models have a complete sourced value.
| Metric | Hy3 Preview | MAI-Thinking-1 | Comparison |
|---|---|---|---|
| Input / output priceUSD per 1M tokens | Hy3 Preview$0 input / $0 output | MAI-Thinking-1Not available | A complete price comparison is not available. |
| Generation speedtokens per second | Hy3 PreviewNot available | MAI-Thinking-1Not available | A complete speed comparison is not available. |
| First-answer latencyseconds to first token | Hy3 PreviewNot available | MAI-Thinking-1Not available | A complete latency comparison is not available. |
| Context windowmaximum listed tokens | Hy3 Preview256K | MAI-Thinking-1256K | Listed context windows are equal. |
Benchmark Deep Dive
AgenticHy3 Preview wins5 benchmarks
CodingMAI-Thinking-1 wins6 benchmarks
Reasoning3 benchmarks
KnowledgeMAI-Thinking-1 wins11 benchmarks
| Benchmark | Hy3 Preview | MAI-Thinking-1 | Result |
|---|---|---|---|
| Artificial Analysis Intelligence IndexSource | 41.2% | — | Not comparable |
| GPQASource | 87.2% | 84.2% | Hy3 Preview leads |
| GPQA-DSource | 87.2% | 84.2% | Hy3 Preview leads |
| HLESource | 25.5% | — | Not comparable |
| AA-Omniscience AccuracySource | 31.5% | — | Not comparable |
| AA-Omniscience Hallucination RateSource | 73.0% | — | Not comparable |
| AA-GPQA DiamondSource | 89.7% | — | Not comparable |
| AA-HLESource | 31.6% | — | Not comparable |
| AA-Omniscience IndexSource | -18.5% | — | Not comparable |
| MMLU-ProSource | — | 85% | Not comparable |
| SimpleQASource | — | 31% | Not comparable |
Math3 benchmarks
Inst. FollowingMAI-Thinking-1 wins1 benchmarks
| Benchmark | Hy3 Preview | MAI-Thinking-1 | Result |
|---|---|---|---|
| IFBenchSource | 63.1% | 85% | MAI-Thinking-1 leads |
Frequently Asked Questions (5)
Which is better, Hy3 Preview or MAI-Thinking-1?
Hy3 Preview 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, Hy3 Preview or MAI-Thinking-1?
MAI-Thinking-1 has the edge for knowledge tasks in this comparison, averaging 72.5 versus 33.8. Inside this category, GPQA is the benchmark that creates the most daylight between them.
Which is better for coding, Hy3 Preview or MAI-Thinking-1?
MAI-Thinking-1 has the edge for coding in this comparison, averaging 65.5 versus 57.8. Inside this category, Terminal-Bench 2.0 is the benchmark that creates the most daylight between them.
Which is better for agentic tasks, Hy3 Preview or MAI-Thinking-1?
Hy3 Preview has the edge for agentic tasks in this comparison, averaging 54.4 versus 46. Inside this category, Terminal-Bench 2.0 is the benchmark that creates the most daylight between them.
Which is better for instruction following, Hy3 Preview or MAI-Thinking-1?
MAI-Thinking-1 has the edge for instruction following in this comparison, averaging 85 versus 63.1. Inside this category, IFBench 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.