Model comparison
GPT-5 Nano vs Hy4 preview
Hy4 preview is the stronger model overall, scoring 45.3 to 33.5 on the Noometry Index. GPT-5 Nano costs 8.2× less per token, which makes it the better buy when Hy4 preview's lead doesn't matter for your workload.
Last verified . 1 shared benchmarks.
TencentSummary
- They share 1 benchmark with published results for both. GPT-5 Nano scores higher in 0 categories and Hy4 preview in 3 categories; 3 gaps are clear of the uncertainty.
- The widest gap is in math, where Hy4 preview leads 55.7 to 29.4.
- The biggest single-benchmark swing is ProofBench: 12% for GPT-5 Nano and 75% for Hy4 preview.
- GPT-5 Nano is cheaper at $0.05 / $0.40 per million input/output tokens, against $0.75 / $2.25 for Hy4 preview.
- Hy4 preview accepts more context: 1.05M tokens versus 400K.
- Hy4 preview has downloadable open weights; the other is API-only.
Side by side
| GPT-5 Nano | Hy4 preview | |
|---|---|---|
| Provider | OpenAI | Tencent |
| Noometry Index | 33.5 | 45.3 |
| Released | 2025-08-07 | 2026-08-28 |
| Weights | Proprietary | Open |
| Context window | 400K | 1.05M |
| Max output | 128K | 64K |
| Input $ / M tokens | $0.05 | $0.75 |
| Output $ / M tokens | $0.40 | $2.25 |
| Results tracked | 49 | 3 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Hy4 preview leads
GPT-5 Nano: 33.6 (#254), Hy4 preview: 51.6 (#38)
| Benchmark | GPT-5 Nano | Hy4 preview |
|---|---|---|
| SWE-bench Verified (bash only) | 34.8% | — |
| LMArena WebDev | — | 1632 |
| WeirdML | 38.1% | — |
| LMArena Coding | 1351 | — |
| ALE-Bench | 718.67 | — |
Agentic & Tool Use Not comparable
GPT-5 Nano: 25.8 (#106), Hy4 preview: —
| Benchmark | GPT-5 Nano | Hy4 preview |
|---|---|---|
| Terminal-Bench | 21.8% | — |
| Berkeley Function Calling Leaderboard | 51.5% | — |
Reasoning Hy4 preview leads
GPT-5 Nano: 16.3 (#306), Hy4 preview: 31.9 (#79)
| Benchmark | GPT-5 Nano | Hy4 preview |
|---|---|---|
| ARC-AGI-2 | 2.6% | — |
| Kagi LLM Benchmark | 62.2% | — |
| NYT Connections (extended) | — | 68.2% |
| ARC-AGI-1 | 20.7% | — |
| Chess Puzzles | 27% | — |
| LMArena Hard Prompts | 1328 | — |
| Mystery Game Puzzles | 9% | — |
| DTBench | 62.7% | — |
| LMCA | 7.9% | — |
| Epoch Capabilities Index | 139.38 | — |
| ForecastBench | 59.1 | — |
Math Hy4 preview leads
GPT-5 Nano: 29.4 (#241), Hy4 preview: 55.7 (#42)
| Benchmark | GPT-5 Nano | Hy4 preview |
|---|---|---|
| ProofBench | 12% | 75% |
| FrontierMath (Tiers 1-3) | 20% | — |
| FrontierMath Tier 4 | 2.4% | — |
| OTIS Mock AIME 2024-2025 | 81.1% | — |
| Omni-MATH | 54.6% | — |
| LMArena Math | 1317 | — |
| MATH Level 5 | 95.2% | — |
| FrontierMath (Feb 2025 set) | 8.3% | — |
| FrontierMath Tier 4 (v1) | 2.1% | — |
Knowledge Not comparable
GPT-5 Nano: 35.9 (#178), Hy4 preview: —
| Benchmark | GPT-5 Nano | Hy4 preview |
|---|---|---|
| GPQA Diamond | 69.4% | — |
| SimpleQA Verified | 11.7% | — |
| MMLU-Pro | 77.8% | — |
| Vectara Hallucination Rate | 10.5% | — |
| GPQA (HELM) | 67.9% | — |
| LMArena Expert | 1321 | — |
Multimodal Not comparable
GPT-5 Nano: 31.3 (#108), Hy4 preview: —
| Benchmark | GPT-5 Nano | Hy4 preview |
|---|---|---|
| LMArena Vision | 1159 | — |
| VPCT | 37.2% | — |
Multilingual Not comparable
GPT-5 Nano: 45.3 (#172), Hy4 preview: —
| Benchmark | GPT-5 Nano | Hy4 preview |
|---|---|---|
| LMArena Non-English | 1313 | — |
| LMArena Chinese | 1356 | — |
| LMArena German | 1327 | — |
| LMArena Japanese | 1226 | — |
| LMArena Korean | 1269 | — |
| LMArena Russian | 1296 | — |
| LMArena Spanish | 1360 | — |
Instruction Following Not comparable
GPT-5 Nano: 75.0 (#79), Hy4 preview: —
| Benchmark | GPT-5 Nano | Hy4 preview |
|---|---|---|
| IFEval | 93.2% | — |
| LMArena Instruction Following | 1306 | — |
Long Context Not comparable
GPT-5 Nano: 31.3 (#281), Hy4 preview: —
| Benchmark | GPT-5 Nano | Hy4 preview |
|---|---|---|
| Fiction.LiveBench | 44.4% | — |
| LMArena Longer Query | 1312 | — |
Writing & Preference Not comparable
GPT-5 Nano: 39.1 (#249), Hy4 preview: —
| Benchmark | GPT-5 Nano | Hy4 preview |
|---|---|---|
| LMArena Text | 1320 | — |
| LMArena Creative Writing | 1249 | — |
| EQ-Bench Creative Writing | 705 | — |
| WildBench | 80.6% | — |
| LMArena Multi-Turn | 1311 | — |
Frequently asked questions
Is GPT-5 Nano better than Hy4 preview?
Hy4 preview is the stronger model overall, scoring 45.3 to 33.5 on the Noometry Index. GPT-5 Nano costs 8.2× less per token, which makes it the better buy when Hy4 preview's lead doesn't matter for your workload.
Which is cheaper, GPT-5 Nano or Hy4 preview?
GPT-5 Nano is cheaper. It lists at $0.05 per million input tokens and $0.40 per million output tokens; Hy4 preview lists at $0.75 and $2.25.
Is GPT-5 Nano or Hy4 preview better for coding?
Hy4 preview scores higher on coding benchmarks: 51.6 versus 33.6 in the Noometry coding category.
Which has the bigger context window?
Hy4 preview does, with 1.05M tokens against 400K.
How many benchmarks do GPT-5 Nano and Hy4 preview share?
1 benchmark has published results for both models. GPT-5 Nano has 49 scored results on Noometry and Hy4 preview has 3.