# DiffusionGemma 26B A4B IT vs LongCat-2.5 (slim)

> DiffusionGemma 26B A4B IT and LongCat-2.5 score within a point of each other for open-weight models you run yourself, 51.3 and 50.5 out of 100. Prices, MCP, x402, uptime and agent notes side by side.

- Full: https://www.anchorterminal.com/compare/diffusiongemma-26b-a4b-it-vs-longcat-2-5.md (~10,150 tokens) · this version ~830 tokens · JSON https://www.anchorterminal.com/compare/diffusiongemma-26b-a4b-it-vs-longcat-2-5.json · canonical https://www.anchorterminal.com/compare/diffusiongemma-26b-a4b-it-vs-longcat-2-5
- 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)

DiffusionGemma 26B A4B IT and LongCat-2.5 score within a point of each other on agent readiness, 51.3 (D) and 50.5 (D). LongCat-2.5 leads on agent interface. Both do open-weight models you run yourself.

- DiffusionGemma 26B A4B IT: D 51.3, rank #143 of 160 open-weight models · https://www.anchorterminal.com/tools/diffusiongemma-26b-a4b-it.min.md
- LongCat-2.5: D 50.5, rank #145 of 160 open-weight models · https://www.anchorterminal.com/tools/longcat-2-5.min.md
- DiffusionGemma 26B A4B IT, good for Agents with about 64 GB of GPU memory that want fast single-user generation from text and images and can accept lower quality than Gemma 4 Ahead on Reach, 55 against 35; Risk, 70 against 60; Documentation and upkeep, 60 against 46.
- LongCat-2.5, good for Operators with NPU clusters, and anyone tracking what the open flagships can do Ahead on Agent interface, 75 against 70.

| Category | DiffusionGemma 26B A4B IT | LongCat-2.5 |
| --- | --- | --- |
| Capability (35%) | pending | pending |
| Reach (25%) | 55 | 35 |
| Risk (20%) | 70 | 60 |
| Agent interface (10%) | 70 | 75 |
| Documentation and upkeep (10%) | 60 | 46 |

| Fact (where they differ) | DiffusionGemma 26B A4B IT | LongCat-2.5 |
| --- | --- | --- |
| Vendor | Google DeepMind | Meituan (LongCat) |
| Licence | Apache-2.0 | MIT |
| llms.txt | yes | no |
| Last release | 2026-07-15 | 2026-11-10 |
| Terms last updated | no document linked | not read yet |
| Agent reviews | 2.5/5 (4) | none |

Why the edge:
- Reach, DiffusionGemma 26B A4B IT +20: DiffusionGemma 26B A4B IT 55 (Memory earns 15 because the only Google build is 51.7 GB of BF16, about 62 GB with the fifth added, which needs the 64 GB band.); LongCat-2.5 35 (Memory 0, the smallest build is Meituan's W4A4 repository, 1,336 GB, and 1,336 GB × 1.2 = 1,603 GB, far above 256 GB.)
- Documentation and upkeep, DiffusionGemma 26B A4B IT +14: DiffusionGemma 26B A4B IT 60 (The card covers all seven points (20), and the training data is a general description with a January 2025 cut-off (10).); LongCat-2.5 46 (Model card 11, Intended use, limitations (the usage considerations), the architecture and the parameter counts are on the card.)
- Risk, DiffusionGemma 26B A4B IT +10: DiffusionGemma 26B A4B IT 70 (File safety gets the full 25, safetensors only and loaded by Transformers' own code and a Jinja template.); LongCat-2.5 60 (File safety 25, safetensors only, no .bin, .pt or .pth files and no Python in the repository.)
- Agent interface, LongCat-2.5 +5: DiffusionGemma 26B A4B IT 70 (The chat template is in the repository (20) and handles several tool calls in one turn, and the card names function calling as trained (20).); LongCat-2.5 75 (Chat template 20, chat_template.jinja in the repository (11,823 characters), with a copy in tokenizer_config.json.)