Model comparison
Hy3 vs Kimi K2.5
Kimi K2.5 is the stronger model overall, scoring 48.1 to 44.2 on the Noometry Index. Hy3 costs 3.9× less per token, which makes it the better buy when Kimi K2.5's lead doesn't matter for your workload.
Last verified . 19 shared benchmarks.
TencentSummary
- They share 19 benchmarks with published results for both. Hy3 scores higher in 0 categories and Kimi K2.5 in 8 categories; 6 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Kimi K2.5 leads 53.6 to 40.8.
- The biggest single-benchmark swing is NYT Connections (extended): 41.2% for Hy3 and 69.9% for Kimi K2.5.
- Hy3 is cheaper at $0.13 / $0.53 per million input/output tokens, against $0.45 / $2.25 for Kimi K2.5.
Side by side
| Hy3 | Kimi K2.5 | |
|---|---|---|
| Provider | Tencent | Moonshot AI |
| Noometry Index | 44.2 | 48.1 |
| Released | 2026-07-06 | 2026-01-27 |
| Weights | Open | Open |
| Context window | 262K | 262K |
| Max output | 128K | 262K |
| Input $ / M tokens | $0.13 | $0.45 |
| Output $ / M tokens | $0.53 | $2.25 |
| Results tracked | 19 | 51 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Kimi K2.5 leads
Hy3: 46.8 (#63), Kimi K2.5: 48.8 (#53)
| Benchmark | Hy3 | Kimi K2.5 |
|---|---|---|
| LMArena WebDev | 1508 | 1437 |
| LMArena Coding | 1464 | 1474 |
| SWE-bench Verified | — | 73.8% |
| SWE-bench Verified (bash only) | — | 70.8% |
| SWE-bench Multilingual | — | 67.3% |
| SciCode | — | 49% |
| WeirdML | — | 45.6% |
| ALE-Bench | — | 821.65 |
Agentic & Tool Use Not comparable
Hy3: —, Kimi K2.5: 34.2 (#48)
| Benchmark | Hy3 | Kimi K2.5 |
|---|---|---|
| Terminal-Bench | — | 43.2% |
| OSWorld | — | 63.3% |
| Vending-Bench 2 | — | 1,198 |
Reasoning Kimi K2.5 leads
Hy3: 26.1 (#136), Kimi K2.5: 31.2 (#80)
| Benchmark | Hy3 | Kimi K2.5 |
|---|---|---|
| NYT Connections (extended) | 41.2% | 69.9% |
| LMArena Hard Prompts | 1447 | 1453 |
| ARC-AGI-2 | — | 11.8% |
| SimpleBench | — | 46.8% |
| Kagi LLM Benchmark | — | 78.5% |
| ARC-AGI-1 | — | 65.3% |
| CritPt | — | 3.1% |
| Chess Puzzles | — | 12% |
| EnigmaEval | — | 3.4% |
| Thematic Generalization | — | 69.4% |
| Epoch Capabilities Index | — | 148.03 |
Math Kimi K2.5 leads
Hy3: 40.1 (#93), Kimi K2.5: 51.8 (#53)
| Benchmark | Hy3 | Kimi K2.5 |
|---|---|---|
| LMArena Math | 1475 | 1470 |
| MathArena Final-Answer Competitions | — | 62.3% |
| OTIS Mock AIME 2024-2025 | — | 92.2% |
| FrontierMath (Feb 2025 set) | — | 27.9% |
| FrontierMath Tier 4 (v1) | — | 4.2% |
Knowledge Kimi K2.5 leads
Hy3: 40.8 (#114), Kimi K2.5: 53.6 (#56)
| Benchmark | Hy3 | Kimi K2.5 |
|---|---|---|
| LMArena Expert | 1460 | 1466 |
| GPQA Diamond | — | 87.6% |
| Humanity's Last Exam | — | 24.4% |
| SimpleQA Verified | — | 34.3% |
| Vectara Hallucination Rate | — | 14.2% |
Multimodal Not comparable
Hy3: —, Kimi K2.5: 41.1 (#39)
| Benchmark | Hy3 | Kimi K2.5 |
|---|---|---|
| LMArena Vision | — | 1269 |
| LMArena Document | — | 1430 |
Multilingual Too close to call
Hy3: 53.5 (#65), Kimi K2.5: 53.9 (#53)
| Benchmark | Hy3 | Kimi K2.5 |
|---|---|---|
| LMArena Non-English | 1426 | 1433 |
| LMArena Chinese | 1493 | 1495 |
| LMArena French | 1461 | 1454 |
| LMArena German | 1439 | 1441 |
| LMArena Japanese | 1392 | 1421 |
| LMArena Korean | 1395 | 1410 |
| LMArena Russian | 1432 | 1435 |
| LMArena Spanish | 1456 | 1450 |
Instruction Following Too close to call
Hy3: 75.1 (#70), Kimi K2.5: 75.3 (#64)
| Benchmark | Hy3 | Kimi K2.5 |
|---|---|---|
| LMArena Instruction Following | 1426 | 1431 |
Long Context Kimi K2.5 leads
Hy3: 44.1 (#75), Kimi K2.5: 52.1 (#7)
| Benchmark | Hy3 | Kimi K2.5 |
|---|---|---|
| LMArena Longer Query | 1442 | 1445 |
| Fiction.LiveBench | — | 86.1% |
| CL-bench | — | 19.3% |
| CL-bench Life | — | 13.2% |
Writing & Preference Kimi K2.5 leads
Hy3: 62.2 (#81), Kimi K2.5: 65.1 (#53)
| Benchmark | Hy3 | Kimi K2.5 |
|---|---|---|
| LMArena Text | 1439 | 1445 |
| LMArena Creative Writing | 1402 | 1423 |
| LMArena Multi-Turn | 1436 | 1444 |
| EQ-Bench Creative Writing | — | 1579 |
Frequently asked questions
Is Hy3 better than Kimi K2.5?
Kimi K2.5 is the stronger model overall, scoring 48.1 to 44.2 on the Noometry Index. Hy3 costs 3.9× less per token, which makes it the better buy when Kimi K2.5's lead doesn't matter for your workload.
Which is cheaper, Hy3 or Kimi K2.5?
Hy3 is cheaper. It lists at $0.13 per million input tokens and $0.53 per million output tokens; Kimi K2.5 lists at $0.45 and $2.25.
Is Hy3 or Kimi K2.5 better for coding?
Kimi K2.5 scores higher on coding benchmarks: 48.8 versus 46.8 in the Noometry coding category.
Which has the bigger context window?
Both accept 262K tokens.
How many benchmarks do Hy3 and Kimi K2.5 share?
19 benchmarks have published results for both models. Hy3 has 19 scored results on Noometry and Kimi K2.5 has 51.