Model comparison
Hy3 vs Hy3 Preview
Head-to-head evidence from 13 shared benchmark results across 4 categories. Overall scores shown here use the public BenchAlign v5 ranking lane.
Sibling matchup inside the Hy3 family.
Public leaderboard positions: Hy3 #80 (Estimated); Hy3 Preview unranked (Not scored). Intervals and evidence labels describe ranking uncertainty, not a guarantee for a specific workload.
Evidence parity. Hy3 and Hy3 Preview share 13 comparable benchmark results. 0 of 8 categories are comparable. 1 result is unique to Hy3; 9 to Hy3 Preview.
Updated July 23, 2026- Shared results
- 13
- Hy3 only
- 1
- Hy3 Preview only
- 9
- Comparable categories
- 0 / 8
Benchmark data for Hy3 and Hy3 Preview is coming soon on BenchLM.
Confidence note. This is a partial-evidence comparison with 13 shared benchmark results across 4 evidence categories; 0 of 8 categories currently have scoreable aggregates for both models. Treat the verdict as directional until coverage is more balanced.
Why this result
BenchLM has partial data for these models, but not enough overlapping benchmark coverage to produce a fair score-level comparison yet.
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 | Δ | Hy3 Preview |
|---|---|---|---|
| Agentic | Hy3Not measured | MarginNo overlap | Hy3 Preview54.4 |
| Coding | Hy3Not measured | MarginNo overlap | Hy3 Preview57.8 |
| Knowledge | Hy3Not measured | MarginNo overlap | Hy3 Preview33.8 |
| Inst. Following | Hy3Not measured | MarginNo overlap | Hy3 Preview63.1 |
Operational comparison
Runtime and commercial metrics are compared only when both models have a complete sourced value.
| Metric | Hy3 | Hy3 Preview | Comparison |
|---|---|---|---|
| Input / output priceUSD per 1M tokens | Hy3$0 input / $0 output | Hy3 Preview$0 input / $0 output | Listed prices are equal. |
| Generation speedtokens per second | Hy3Not available | Hy3 PreviewNot available | A complete speed comparison is not available. |
| First-answer latencyseconds to first token | Hy3Not available | Hy3 PreviewNot available | A complete latency comparison is not available. |
| Context windowmaximum listed tokens | Hy3256K | Hy3 Preview256K | Listed context windows are equal. |
Benchmark Deep Dive
Agentic5 benchmarks
Coding5 benchmarks
Reasoning2 benchmarks
Knowledge9 benchmarks
| Benchmark | Hy3 | Hy3 Preview | Result |
|---|---|---|---|
| Artificial Analysis Intelligence IndexSource | 41.2% | 41.2% | Tie |
| AA-GPQA DiamondSource | 89.7% | 89.7% | Tie |
| AA-HLESource | 31.6% | 31.6% | Tie |
| AA-Omniscience IndexSource | -18.5% | -18.5% | Tie |
| AA-Omniscience AccuracySource | 31.5% | 31.5% | Tie |
| AA-Omniscience Hallucination RateSource | 73.0% | 73.0% | Tie |
| GPQASource | — | 87.2% | Not comparable |
| GPQA-DSource | — | 87.2% | Not comparable |
| HLESource | — | 25.5% | Not comparable |
Multimodal1 benchmarks
| Benchmark | Hy3 | Hy3 Preview | Result |
|---|---|---|---|
| Design Arena WebsiteSource | 1229 | — | Not comparable |
Inst. Following1 benchmarks
| Benchmark | Hy3 | Hy3 Preview | Result |
|---|---|---|---|
| IFBenchSource | — | 63.1% | Not comparable |
Frequently Asked Questions (3)
Can I compare Hy3 and Hy3 Preview on BenchLM yet?
Not fully yet. BenchLM is tracking both models, but the sourced benchmark breakdown for this comparison is still coming soon.
Why does this comparison show “coming soon”?
BenchLM only shows category winners and benchmark-level calls when we have sourced results that can be compared fairly. For these models, the public benchmark coverage is not complete enough yet.
What data is available for Hy3 and Hy3 Preview today?
Hy3: $0.00 input / $0.00 output per 1M tokens Hy3 Preview: $0.00 input / $0.00 output per 1M tokens Both model pages still include creator, context window, reasoning mode, and other metadata while benchmark coverage fills in.
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.