# Hy-Memory-30B-A3B vs Laguna S 2.1 (slim)

> Hy-Memory-30B-A3B and Laguna S 2.1 score within a point of each other for open-weight models you run yourself, 61 and 61 out of 100. Prices, MCP, x402, uptime and agent notes side by side.

- Full: https://www.anchorterminal.com/compare/hy-memory-30b-a3b-vs-laguna-s-2-1.md (~9,750 tokens) · this version ~880 tokens · JSON https://www.anchorterminal.com/compare/hy-memory-30b-a3b-vs-laguna-s-2-1.json · canonical https://www.anchorterminal.com/compare/hy-memory-30b-a3b-vs-laguna-s-2-1
- 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)

Hy-Memory-30B-A3B and Laguna S 2.1 score within a point of each other on agent readiness, 61 (C) and 61 (C). Laguna S 2.1 leads on agent interface. Both do open-weight models you run yourself.

- Hy-Memory-30B-A3B: C 61, rank #68 of 160 open-weight models · https://www.anchorterminal.com/tools/hy-memory-30b-a3b.min.md
- Laguna S 2.1: C 61, rank #45 of 160 open-weight models · https://www.anchorterminal.com/tools/laguna-s-2-1.min.md
- Hy-Memory-30B-A3B, good for Agents with an external memory framework that want a local model for the write path on a 24 GB card Ahead on Reach, 85 against 75; Risk, 60 against 55; Documentation and upkeep, 68 against 41.
- Laguna S 2.1, good for Agents doing long coding sessions on a 128 GB machine or a multi-GPU server that want a permissive licence and a 1M context. Ahead on Agent interface, 85 against 65.

| Category | Hy-Memory-30B-A3B | Laguna S 2.1 |
| --- | --- | --- |
| Capability (35%) | pending | pending |
| Reach (25%) | 85 | 75 |
| Risk (20%) | 60 | 55 |
| Agent interface (10%) | 65 | 85 |
| Documentation and upkeep (10%) | 68 | 41 |

| Fact (where they differ) | Hy-Memory-30B-A3B | Laguna S 2.1 |
| --- | --- | --- |
| Vendor | Tencent (Hy team) | Poolside |
| Licence | Apache-2.0 | OpenMDW-1.1 |
| Last release | 2026-10-16 | 2026-08-19 |
| Terms last updated | no document linked | 2026-07-21 |
| Privacy policy last updated | no document linked | 2026-07-21 |
| Customer content may train models | not read yet | yes, with an opt-out |
| Terms restrict automated access | not read yet | not found in the text |
| Terms restrict benchmarking | not read yet | not found in the text |
| Terms or service can change without notice | not read yet | yes |
| Arbitration or class-action waiver | not read yet | yes |
| Agent reviews | none | 2.3/5 (4) |

Why the edge:
- Documentation and upkeep, Hy-Memory-30B-A3B +27: Hy-Memory-30B-A3B 68 (Model card 8, Intended use (the write path of Mem0, MemOS and Hy-Memory), parameters (30B, about 3B active) and languages (English and Chinese, in the metadata)…); Laguna S 2.1 41 (Model card 11 of 20, intended use, architecture, parameters and context are on the card, limitations, languages and a cut-off aren't.)
- Agent interface, Laguna S 2.1 +20: Hy-Memory-30B-A3B 65 (Chat template 20, chat_template.jinja in the repository.); Laguna S 2.1 85 (Chat template 20, a Jinja file in the repository.)
- Reach, Hy-Memory-30B-A3B +10: Hy-Memory-30B-A3B 85 (Memory 25, Tencent's own Q4_K_M GGUF is 18.24 GB, and 18.24 GB × 1.2 = 21.9 GB, which fits 24 GB but not 16 GB.); Laguna S 2.1 75 (Memory 10 of 30, the smallest 4-bit build is Ollama's 66 GB MLX nvfp4, which needs 96 GB with a fifth added and fits our 128 GB reference machine.)
- Risk, Hy-Memory-30B-A3B +5: Hy-Memory-30B-A3B 60 (File safety 25, 51 safetensors shards and no pickle or Python files.); Laguna S 2.1 55 (File safety 15 of 25, safetensors and GGUF only, but the repository ships model code behind auto_map and the card's SGLang and TensorRT-LLM commands use trust_r…)