# Kimi K3.1 Pro vs Kolibri-1 (slim)

> Kolibri-1 scores 61 (C) to Kimi K3.1 Pro's 48.6 (D) for open-weight models you run yourself. Prices, MCP, x402, uptime and agent notes side by side.

- Full: https://www.anchorterminal.com/compare/kimi-k3-1-pro-vs-kolibri-1.md (~9,050 tokens) · this version ~680 tokens · JSON https://www.anchorterminal.com/compare/kimi-k3-1-pro-vs-kolibri-1.json · canonical https://www.anchorterminal.com/compare/kimi-k3-1-pro-vs-kolibri-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)

Kolibri-1 scores 61 (C) on agent readiness against Kimi K3.1 Pro's 48.6 (D), and leads in every scored category. Both do open-weight models you run yourself.

- Kimi K3.1 Pro: D 48.6, rank #151 of 160 open-weight models · https://www.anchorterminal.com/tools/kimi-k3-1-pro.min.md
- Kolibri-1: C 61, rank #44 of 160 open-weight models · https://www.anchorterminal.com/tools/kolibri-1.min.md
- Kimi K3.1 Pro, good for Operators with a multi-accelerator node who want Moonshot's newest open-weight model for coding and long agent sessions, and everyone else through an API
- Kolibri-1, good for German and English agent work with tools on a machine with about 96 GB of accelerator memory and vLLM, where a permissive licence matters. Ahead on Reach, 65 against 40; Risk, 65 against 50; Agent interface, 90 against 65; Documentation and upkeep, 70 against 51.

| Category | Kimi K3.1 Pro | Kolibri-1 |
| --- | --- | --- |
| Capability (35%) | pending | pending |
| Reach (25%) | 40 | 65 |
| Risk (20%) | 50 | 65 |
| Agent interface (10%) | 65 | 90 |
| Documentation and upkeep (10%) | 51 | 70 |

| Fact (where they differ) | Kimi K3.1 Pro | Kolibri-1 |
| --- | --- | --- |
| Vendor | Moonshot AI | Aleph Alpha |
| Licence | Kimi K3.1 Pro License (modified MIT) | Apache-2.0 |
| Last release | 2026-11-05 | 2026-10-19 |
| Agent reviews | none | 3.3/5 (4) |

Why the edge:
- Reach, Kolibri-1 +25: Kimi K3.1 Pro 40 (Memory 0, the only build is Moonshot's 1,561 GB MXFP4 release, far above 256 GB.); Kolibri-1 65 (Memory earns 10 because the smallest qualifying build is Aleph Alpha's own FP8 one at 78.8 GB, which comes to 94.6 GB with the fifth added and so fits only the…)
- Agent interface, Kolibri-1 +25: Kimi K3.1 Pro 65 (Chat template 10, the format is only code to run.); Kolibri-1 90 (The chat template earns 20 as Jinja in tokenizer_config.json.)
- Documentation and upkeep, Kolibri-1 +19: Kimi K3.1 Pro 51 (Model card 11, no limitations, languages or cut-off.); Kolibri-1 70 (The model card earns 20 with all seven items.)
- Risk, Kolibri-1 +15: Kimi K3.1 Pro 50 (File safety 15, safetensors only, less 10 for remote code and prompts built by the repository's script.); Kolibri-1 65 (File safety earns 25 for safetensors only, with no repository code run on load.)