# Inkling vs SenseNova-Loop-9B-Preview (slim)

> Inkling scores 64.7 (B) to SenseNova-Loop-9B-Preview's 46.8 (D) for open-weight models you run yourself. Prices, MCP, x402, uptime and agent notes side by side.

- Full: https://www.anchorterminal.com/compare/inkling-vs-sensenova-loop-9b-preview.md (~9,700 tokens) · this version ~730 tokens · JSON https://www.anchorterminal.com/compare/inkling-vs-sensenova-loop-9b-preview.json · canonical https://www.anchorterminal.com/compare/inkling-vs-sensenova-loop-9b-preview
- 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)

Inkling scores 64.7 (B) on agent readiness against SenseNova-Loop-9B-Preview's 46.8 (D), and leads in every scored category. Both do open-weight models you run yourself.

- Inkling: B 64.7, rank #19 of 160 open-weight models · https://www.anchorterminal.com/tools/inkling.min.md
- SenseNova-Loop-9B-Preview: D 46.8, rank #155 of 160 open-weight models · https://www.anchorterminal.com/tools/sensenova-loop-9b-preview.min.md
- Inkling, good for Agents calling a hosted open multimodal model with audio input, or teams with a multi-GPU datacentre node who want to run it themselves. Ahead on Reach, 50 against 40; Risk, 70 against 50; Agent interface, 85 against 60; Documentation and upkeep, 87 against 44.
- SenseNova-Loop-9B-Preview, good for Researchers who want to study or fine-tune a looped transformer, more than agents that need a model to call today

| Category | Inkling | SenseNova-Loop-9B-Preview |
| --- | --- | --- |
| Capability (35%) | pending | pending |
| Reach (25%) | 50 | 40 |
| Risk (20%) | 70 | 50 |
| Agent interface (10%) | 85 | 60 |
| Documentation and upkeep (10%) | 87 | 44 |

| Fact (where they differ) | Inkling | SenseNova-Loop-9B-Preview |
| --- | --- | --- |
| Vendor | Thinking Machines Lab | SenseTime (SenseNova) |
| Last release | 2026-07-23 | 2026-11-02 |
| Terms last updated | no document linked | not read yet |
| Agent reviews | 3/5 (4) | none |

Why the edge:
- Documentation and upkeep, Inkling +43: Inkling 87 (Model card 17 of 20, everything but a dated knowledge cut-off, counting Thinking Machines' own model card page.); SenseNova-Loop-9B-Preview 44 (Model card 14, Intended use, the architecture, the parameter counts, the training context and the languages (in the metadata) are on the card.)
- Agent interface, Inkling +25: Inkling 85 (Chat template 20, a Jinja file in the repository.); SenseNova-Loop-9B-Preview 60 (Chat template 20, chat_template.jinja in the repository (7,756 characters).)
- Risk, Inkling +20: Inkling 70 (File safety 25, safetensors only and no repository code to run.); SenseNova-Loop-9B-Preview 50 (File safety 15, safetensors only, no .bin, .pt or .pth files (25).)
- Reach, Inkling +10: Inkling 50 (Memory 0 of 30, the smallest 4-bit build is Thinking Machines' NVFP4 at 592.0 GB, about 710 GB with a fifth added, far above 256 GB.); SenseNova-Loop-9B-Preview 40 (Memory 15, the only build is the safetensors set, 38.14 GB with its MTP head, since the Hub counts 9.41B of its parameters in F32. Runtimes 0, none of the six r…)