Model comparison
Hy3 vs Llama 3.1-70B
Hy3 is the stronger model overall, scoring 44.2 to 29.6 on the Noometry Index.
Last verified . 17 shared benchmarks.
Tencent
MetaSummary
- They share 17 benchmarks with published results for both. Hy3 scores higher in 8 categories and Llama 3.1-70B in 0 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where Hy3 leads 62.2 to 35.4.
- Hy3 is cheaper at $0.0825 / $0.33 per million input/output tokens, against $0.40 / $0.40 for Llama 3.1-70B.
- Hy3 accepts more context: 262K tokens versus 128K.
Side by side
| Hy3 | Llama 3.1-70B | |
|---|---|---|
| Provider | Tencent | Meta |
| Noometry Index | 44.2 | 29.6 |
| Released | 2026-07-06 | 2024-07-23 |
| Weights | Open | Open |
| Context window | 262K | 128K |
| Max output | 128K | 4K |
| Input $ / M tokens | $0.0825 | $0.40 |
| Output $ / M tokens | $0.33 | $0.40 |
| Results tracked | 19 | 35 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Hy3 leads
Hy3: 46.8 (#63), Llama 3.1-70B: 30.3 (#296)
| Benchmark | Hy3 | Llama 3.1-70B |
|---|---|---|
| LMArena Coding | 1464 | 1260 |
| LMArena WebDev | 1508 | — |
| WeirdML | — | 9% |
| BigCodeBench Instruct | — | 46.1% |
| BigCodeBench Complete | — | 54.8% |
Agentic & Tool Use Not comparable
Hy3: —, Llama 3.1-70B: 25.1 (#112)
| Benchmark | Hy3 | Llama 3.1-70B |
|---|---|---|
| TheAgentCompany | — | 6.9% |
| BALROG | — | 27.9% |
Reasoning Hy3 leads
Hy3: 26.1 (#136), Llama 3.1-70B: 21.6 (#220)
| Benchmark | Hy3 | Llama 3.1-70B |
|---|---|---|
| LMArena Hard Prompts | 1447 | 1241 |
| NYT Connections (extended) | 41.2% | — |
| DTBench | — | 60% |
| LMCA | — | 14.8% |
| Epoch Capabilities Index | — | 125.92 |
Math Hy3 leads
Hy3: 40.1 (#93), Llama 3.1-70B: 13.5 (#304)
| Benchmark | Hy3 | Llama 3.1-70B |
|---|---|---|
| LMArena Math | 1475 | 1252 |
| OTIS Mock AIME 2024-2025 | — | 3.6% |
| Omni-MATH | — | 21% |
| MATH Level 5 | — | 36.7% |
Knowledge Hy3 leads
Hy3: 40.8 (#114), Llama 3.1-70B: 24.2 (#269)
| Benchmark | Hy3 | Llama 3.1-70B |
|---|---|---|
| LMArena Expert | 1460 | 1209 |
| GPQA Diamond | — | 44.2% |
| MMLU-Pro | — | 65.3% |
| GPQA (HELM) | — | 42.6% |
| MMLU | — | 80.1% |
Multilingual Hy3 leads
Hy3: 53.5 (#65), Llama 3.1-70B: 38.8 (#225)
| Benchmark | Hy3 | Llama 3.1-70B |
|---|---|---|
| LMArena Non-English | 1426 | 1219 |
| LMArena Chinese | 1493 | 1215 |
| LMArena French | 1461 | 1261 |
| LMArena German | 1439 | 1222 |
| LMArena Japanese | 1392 | 1132 |
| LMArena Korean | 1395 | 1140 |
| LMArena Russian | 1432 | 1234 |
| LMArena Spanish | 1456 | 1253 |
Instruction Following Hy3 leads
Hy3: 75.1 (#70), Llama 3.1-70B: 65.3 (#223)
| Benchmark | Hy3 | Llama 3.1-70B |
|---|---|---|
| LMArena Instruction Following | 1426 | 1231 |
| IFEval | — | 82.1% |
Long Context Hy3 leads
Hy3: 44.1 (#75), Llama 3.1-70B: 37.6 (#214)
| Benchmark | Hy3 | Llama 3.1-70B |
|---|---|---|
| LMArena Longer Query | 1442 | 1241 |
Writing & Preference Hy3 leads
Hy3: 62.2 (#81), Llama 3.1-70B: 35.4 (#267)
| Benchmark | Hy3 | Llama 3.1-70B |
|---|---|---|
| LMArena Text | 1439 | 1261 |
| LMArena Creative Writing | 1402 | 1232 |
| LMArena Multi-Turn | 1436 | 1256 |
| EQ-Bench Creative Writing | — | 784 |
| WildBench | — | 75.8% |
Frequently asked questions
Is Hy3 better than Llama 3.1-70B?
Hy3 is the stronger model overall, scoring 44.2 to 29.6 on the Noometry Index.
Which is cheaper, Hy3 or Llama 3.1-70B?
Hy3 is cheaper. It lists at $0.0825 per million input tokens and $0.33 per million output tokens; Llama 3.1-70B lists at $0.40 and $0.40.
Is Hy3 or Llama 3.1-70B better for coding?
Hy3 scores higher on coding benchmarks: 46.8 versus 30.3 in the Noometry coding category.
Which has the bigger context window?
Hy3 does, with 262K tokens against 128K.
How many benchmarks do Hy3 and Llama 3.1-70B share?
17 benchmarks have published results for both models. Hy3 has 19 scored results on Noometry and Llama 3.1-70B has 35.