[Integrations](https://spiral.ai/integrations) › [Windsurf](https://spiral.ai/integrations/windsurf) › [Ollama model layer metadata](https://spiral.ai/integrations/windsurf/ollama-model-layer-metadata)

# How to Use Ollama model layer metadata in Windsurf

Returns a model layer's architecture, context length, embedding dimensions, and tensor metadata. Use it to check the actual build before choosing an embedding model or comparing quantization.

|  |  |
| --- | --- |
| Model | Ollama model layer metadata |
| Available through | Alexandria |
| Category | Developer |
| Price | $0.00873 per call. |
| Returns | json |
| Runs | In the background |
| Setup | About 2 minutes, no API key |

## Prompts to try with Ollama model layer metadata in Windsurf

- Use Ollama model layer metadata to inspect qwen3-embedding:0.6b, then compare its embedding dimension and context length with the assumptions in this project's retrieval code.
- Use Ollama model layer metadata to look this up and give me the result as a table. Save the result in ./assets.
- Pull the latest numbers from Ollama model layer metadata and tell me what changed since last week.
- Check Ollama model layer metadata and summarize what you find in three bullet points.
- Pull the newest data from Ollama model layer metadata and give me the result as a table.

## How to set up Ollama model layer metadata in Windsurf

Pick the way that suits you. Each takes about two minutes.

### Use Spiral's MCP server

1. **Add the server**: Open the Customize page from the Agents window and click MCPs. Under This computer, open the config file, or edit it directly: `~/.config/devin/mcp_config.json` on macOS and Linux, `%APPDATA%\devin\mcp_config.json` on Windows. Add this:

```json
{
  "mcpServers": {
    "spiral": {
      "url": "https://api.spiral.ai/mcp"
    }
  }
}
```

2. **Sign in**: Click Authenticate on the Spiral server in Customize › MCPs, or let the agent call Spiral once: your browser opens. Sign in to Spiral, pick the organization and click Authorize. No API key.

```bash
devin mcp login spiral
```

3. **Check it**: Spiral is listed under This computer in Customize › MCPs. From the terminal:

```bash
devin mcp list
```

### Use the Spiral skill

1. **Send Windsurf this message**: Your agent reads Spiral's skill and sets itself up.

```
set up https://spiral.ai/SKILL.md
```

2. **Sign in or paste a key**: Approve Spiral in the browser, or give the agent a key from the console.

### Use the Spiral CLI

1. **Install and sign in**: Installs the Spiral CLI, signs you in and adds its skill to your coding agents.

```bash
npx -y spiral-cli@latest init
```

### Call the Spiral API

1. **Create an API key**: In the console, under API keys.

2. **Run Ollama model layer metadata**: One request; the result comes back by polling or a webhook.

```bash
curl https://api.spiral.ai/v1/runs \
  -H "Authorization: Bearer $SPIRAL_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: $(uuidgen)" \
  -d '{
    "provider": "alexandria",
    "model": "ollama-com/models/layer_metadata",
    "input": {
      "tag": "<tag>",
      "model": "<model>",
      "layer_digest": "<layer_digest>"
    }
  }'
```

Then paste one of the prompts above: Windsurf finds Ollama model layer metadata on Spiral, shows the price and runs it.

Setup guide in the docs: https://spiral.ai/docs/quickstarts/windsurf

## Ollama model layer metadata tools and prices

- [Ollama model library · Layer metadata](https://spiral.ai/tools/alexandria/ollama-com/models/layer_metadata) (Alexandria): $0.00873 per call.

## Settings you can ask for

| Setting | Allowed values | Default | Description |
| --- | --- | --- | --- |
| include_tensors |  | — | Return the tensor table (name, type, shape); a few hundred rows. Default false. |

## What can you do with Ollama model layer metadata in Windsurf?

Build a model selection report that connects your project's requirements to Ollama descriptions, tags, and layer metadata. A coding agent can compare those results with the embedding dimensions or model references already in your code.

### Choose an embedding model for document search

Find official and community candidates, review a publisher's uploads, and read model descriptions. Compare tags, download contents, context lengths, and embedding dimensions before choosing a build for a retrieval trial.

### Prepare a coding-model evaluation

Collect candidates from the library, search results, and selected publishers. Read their stated tasks, compare available builds, and inspect tag contents and tensor types to define which exact models to benchmark.

### Plan model downloads for an offline deployment

Build a shortlist from official listings and public publisher catalogs, then review model descriptions and tags. Use manifest sizes and layer metadata to document the selected downloads, architecture, and quantization.

### Write a technical model comparison

Compare models discovered through the official library, search, and publisher pages. Support the report with README descriptions, tag choices, download sizes, layer identifiers, and metadata from the selected builds.

### Check an embedding migration

Compare nomic-embed-text and qwen3-embedding builds using their model descriptions, tags, layers, context lengths, and embedding dimensions. Identify differences to account for before testing a replacement search index.

### Inspect mixed tensor quantization

Choose a llama3.2 tag, identify its model layer, and examine individual tensor types and shapes. Explain where the tensor table contains types such as Q4_K, Q6_K, and F32.

## Overview

- [Alexandria](https://spiral.ai/providers/alexandria)
- [All Windsurf guides](https://spiral.ai/integrations/windsurf)

## Ollama model layer metadata in other agents

- [Amp](https://spiral.ai/integrations/amp/ollama-model-layer-metadata)
- [ChatGPT](https://spiral.ai/integrations/chatgpt/ollama-model-layer-metadata)
- [Claude](https://spiral.ai/integrations/claude-desktop/ollama-model-layer-metadata)
- [Claude Code](https://spiral.ai/integrations/claude-code/ollama-model-layer-metadata)
- [Cline](https://spiral.ai/integrations/cline/ollama-model-layer-metadata)
- [Codex](https://spiral.ai/integrations/codex/ollama-model-layer-metadata)
- [Cursor](https://spiral.ai/integrations/cursor/ollama-model-layer-metadata)
- [Devin](https://spiral.ai/integrations/devin/ollama-model-layer-metadata)
- [dots](https://spiral.ai/integrations/dots/ollama-model-layer-metadata)
- [Factory Droid](https://spiral.ai/integrations/factory-droid/ollama-model-layer-metadata)
- [Gemini](https://spiral.ai/integrations/gemini/ollama-model-layer-metadata)
- [Gemini CLI](https://spiral.ai/integrations/gemini-cli/ollama-model-layer-metadata)
- [Google Antigravity](https://spiral.ai/integrations/antigravity/ollama-model-layer-metadata)
- [goose](https://spiral.ai/integrations/goose/ollama-model-layer-metadata)
- [Grok](https://spiral.ai/integrations/grok/ollama-model-layer-metadata)
- [Grok Build](https://spiral.ai/integrations/grok-build/ollama-model-layer-metadata)
- [Hermes](https://spiral.ai/integrations/hermes/ollama-model-layer-metadata)
- [Instinct](https://spiral.ai/integrations/instinct/ollama-model-layer-metadata)
- [Manus](https://spiral.ai/integrations/manus/ollama-model-layer-metadata)
- [Muse](https://spiral.ai/integrations/muse/ollama-model-layer-metadata)
- [Muse Code](https://spiral.ai/integrations/muse-code/ollama-model-layer-metadata)
- [n8n](https://spiral.ai/integrations/n8n/ollama-model-layer-metadata)
- [Notion](https://spiral.ai/integrations/notion/ollama-model-layer-metadata)
- [OpenClaw](https://spiral.ai/integrations/openclaw/ollama-model-layer-metadata)
- [opencode](https://spiral.ai/integrations/opencode/ollama-model-layer-metadata)
- [Perplexity](https://spiral.ai/integrations/perplexity/ollama-model-layer-metadata)
- [Raycast](https://spiral.ai/integrations/raycast/ollama-model-layer-metadata)
- [VS Code](https://spiral.ai/integrations/vscode/ollama-model-layer-metadata)
- [Zapier](https://spiral.ai/integrations/zapier/ollama-model-layer-metadata)

## More Windsurf guides

- [Ollama model details](https://spiral.ai/integrations/windsurf/ollama-model-details)
- [Ollama model search](https://spiral.ai/integrations/windsurf/ollama-model-search)
- [OSV.dev vulnerability details](https://spiral.ai/integrations/windsurf/osv-dev-vulnerability-details)
- [Ollama model tag details](https://spiral.ai/integrations/windsurf/ollama-model-tag-details)
- [OSV.dev package vulnerability lookup](https://spiral.ai/integrations/windsurf/osv-dev-package-vulnerability-lookup)
- [Ollama model tags](https://spiral.ai/integrations/windsurf/ollama-model-tags)

## Questions about Ollama model layer metadata in Windsurf

**Can Windsurf use Ollama model layer metadata?**

Yes. Add Spiral to Windsurf as an MCP server, and Windsurf can find and run Ollama model layer metadata like any built-in tool.

**How much does Ollama model layer metadata cost in Windsurf?**

$0.00873 per call. You pay from your Spiral balance as you go, with no subscription, and a failed run is not charged. Your Windsurf plan is separate.

**Do I need an Alexandria account to use Ollama model layer metadata in Windsurf?**

No. Spiral runs Ollama model layer metadata on its own Alexandria account. You sign in to Spiral once and never paste a key.

**How long does it take to set up Ollama model layer metadata in Windsurf?**

About two minutes: add the server, then sign in, then check it.

**Does Windsurf wait while Ollama model layer metadata runs?**

Ollama model layer metadata runs in the background. Windsurf starts the run, checks on it and gives you the json when it's ready.

**Does Ollama include community uploads?**

Yes. Ollama has public publisher namespaces as well as its official library. For example, aiconjured's catalog includes Qwen3-Embedding-0.6B-Q8-NVFP4.

**How many builds can one model have?**

The count depends on the model. The inspected llama3.2 page lists 63 tags across choices that include 1B and 3B models, so a model name alone does not identify one download.

**How does Ollama show how recently a model changed?**

Catalog and tag entries display relative update ages. For example, the inspected llama3.2:1b tag shows “2 years ago” beside its size and context information.

**Where can I check an embedding build's context length and dimensions?**

Ollama's model layer metadata provides both. The inspected nomic-embed-text layer reports a context length of 2,048 and an embedding length of 768.

**Can every model in Ollama's catalog answer chat questions?**

No. For example, nomic-embed-text is explicitly limited to generating embeddings, so use its catalog data to evaluate search or retrieval models.

