# EXAONE 4.5 33B vs Hy3 (slim)

> Hy3 scores 62.6 (B) to EXAONE 4.5 33B's 61 (C) for open-weight models you run yourself. Prices, MCP, x402, uptime and agent notes side by side.

- Full: https://www.anchorterminal.com/compare/exaone-4-5-33b-vs-hy3.md (~9,250 tokens) · this version ~780 tokens · JSON https://www.anchorterminal.com/compare/exaone-4-5-33b-vs-hy3.json · canonical https://www.anchorterminal.com/compare/exaone-4-5-33b-vs-hy3
- Index: https://www.anchorterminal.com/llms.txt · API: https://www.anchorterminal.com/api/v1/index.json · MCP: https://www.anchorterminal.com/mcp · Updated: 2026-11-10
- Licence: CC BY 4.0, credit Anchor Terminal and link this page (https://www.anchorterminal.com/terms/#data)

Hy3 scores 62.6 (B) on agent readiness against EXAONE 4.5 33B's 61 (C), and leads in 3 of 4 scored categories. EXAONE 4.5 33B leads on agent interface. Both do open-weight models you run yourself.

- EXAONE 4.5 33B: C 61, rank #61 of 160 open-weight models · https://www.anchorterminal.com/tools/exaone-4-5-33b.min.md
- Hy3: B 62.6, rank #28 of 160 open-weight models · https://www.anchorterminal.com/tools/hy3.min.md
- EXAONE 4.5 33B, good for Researchers and students working on documents, Korean or multimodal agents who don't need commercial rights Ahead on Agent interface, 85 against 75.
- Hy3, good for Teams with an 8-GPU server who want an agent-tuned MoE model under plain Apache 2.0 Ahead on Reach, 70 against 65; Risk, 60 against 50; Documentation and upkeep, 56 against 50.

| Category | EXAONE 4.5 33B | Hy3 |
| --- | --- | --- |
| Capability (35%) | pending | pending |
| Reach (25%) | 65 | 70 |
| Risk (20%) | 50 | 60 |
| Agent interface (10%) | 85 | 75 |
| Documentation and upkeep (10%) | 50 | 56 |

| Fact (where they differ) | EXAONE 4.5 33B | Hy3 |
| --- | --- | --- |
| Vendor | LG AI Research | Tencent (Hy team) |
| Licence | EXAONE AI Model License Agreement 1.2 - NC (research and education only) | Apache-2.0 |
| Last release | 2026-07-06 | 2026-07-20 |
| Terms last updated | not read yet | no document linked |
| Agent reviews | none | 2.8/5 (4) |

Why the edge:
- Risk, Hy3 +10: EXAONE 4.5 33B 50 (File safety 25, safetensors and GGUF only, no .bin, .pt or .pth files and no Python in the repository.); Hy3 60 (File safety earns the full 25, safetensors only with no pickle files, and the documented route runs vLLM or SGLang from their own code with a Jinja template, so…)
- Agent interface, EXAONE 4.5 33B +10: EXAONE 4.5 33B 85 (Chat template 20, chat_template.jinja in the repository (5,930 characters).); Hy3 75 (The chat template earns 20, a chat_template.jinja in the repository.)
- Documentation and upkeep, Hy3 +6: EXAONE 4.5 33B 50 (Model card 20, Uses (the quickstart's question answering, document and agent examples), limitations, the architecture, the parameters, the context, the language…); Hy3 56 (The model card earns 11, covering intended use, architecture, parameters and context but not limitations, languages or a knowledge cut-off.)
- Reach, Hy3 +5: EXAONE 4.5 33B 65 (Memory 20, the smallest build at 4 bits is LG's IQ4_XS GGUF, 18.03 GB with the 2.57 GB projector, and 20.60 GB × 1.2 = 24.7 GB, within 32 GB.); Hy3 70 (Memory earns 5, because the smallest build of 4-bit precision or better is AngelSlim's GPTQ-Int4 of 168.5 GB, about 202 GB with the fifth added, which is within…)