Hy3 vs Kimi K3: Specs & Benchmark Comparison
Hy3 is developed by Tencent, while Kimi K3 comes from Moonshot AI. Both were released in July 2026. Kimi K3 is the larger model at roughly 2.8 trillion parameters, against 295 billion for Hy3.
The two models share 8 published benchmarks. Kimi K3 leads on 8 of them. The widest gaps are on DeepSWE, where Kimi K3 scores 67.5% against 28.0%; Toolathlon, where Kimi K3 scores 73.2% against 48.5%. Averaged across everything we track, Hy3 sits at 57.1% and Kimi K3 at 67.8%.
Kimi K3 is available through an API at $3 per million input tokens with a 1M-token context window. We do not track a hosted API for Hy3, so pricing and context cannot be compared directly.
| Characteristic | Hy3 | Kimi K3 |
|---|---|---|
| Company | Tencent | Moonshot AI |
| Release Date | July 6, 2026 | July 16, 2026 |
| Parameters | 295B | 2.8T |
| Multimodal | No | Yes |
| Context (input) | — | 1.0M |
| Context (output) | — | 1.0M |
| Input Price / 1M | — | $3.00 |
| Output Price / 1M | — | $15.00 |
| Average Score | 57.1% | 67.8% |
| Benchmarks | ||
| DeepSWE | 28.0% | 67.5% |
| Toolathlon | 48.5% | 73.2% |
| Terminal-Bench 2.1 | 71.7% | 88.3% |
| APEX-Agents | 25.6% | 37.6% |
| BrowseComp | 84.2% | 91.2% |
| MCP Atlas | 79.1% | 84.2% |
| DeepSearchQA | 91.0% | 95.0% |
| GPQA | 90.4% | 94.0% |
Visual Benchmark Comparison
Verdict
Both models show equal results — the choice depends on your specific use case.
Both models show comparable average scores: Hy3 — 0.6, Kimi K3 — 0.7.
Both models were released around the same time: 7/6/2026 and 7/16/2026.
More About These Models
Related Comparisons
Frequently Asked Questions
Which is better for coding — Hy3 or Kimi K3?
Which model is cheaper — Hy3 or Kimi K3?
Which has a larger context window — Hy3 or Kimi K3?
The Hy3 and Kimi K3 comparison is updated for 2026. Data includes benchmark results, API pricing, context window size and other specifications. For more detailed information, visit the Hy3 or Kimi K3 page. See also the complete list of AI model comparisons.