Spiral gives your agent every tool with one account. Browse the tools →

Codex · LLMReleased 2026-09-12

How to Use Schematron V2 Turbo in Codex

Schematron V2 Turbo turns HTML into JSON for high-volume extraction tasks. It prioritizes throughput, while Schematron V2 Small prioritizes extraction quality on complex schemas and long pages.

Prompts to try with Schematron V2 Turbo in Codex

Copy an example into Codex, or ask in your own words.

Use Schematron V2 Turbo to extract product names and specifications from the project's saved product-page HTML. Derive a JSON Schema from our product type, put extraction rules in its descriptions, and use temperature 0.

Schematron V2 TurboCodex
Try asking
  • Use Schematron V2 Turbo on our saved job-page HTML. Define a schema for title, employer, location, and application URL, then compare the extracted records with the expected fixtures.

  • Use Schematron V2 Turbo to extract article headlines, bylines, and publication dates from our HTML fixtures. Describe publication date precisely in the schema so it is distinguished from update dates.

  • Use Schematron V2 Turbo to process the project's event-page HTML into event records. Clean scripts and styles first, preserve displayed date text, and split the work so each response stays below 8,192 tokens.

  • Use Schematron V2 Turbo to extract company names, locations, and website links from our directory-page samples. Put the rules in the JSON Schema and report records that fail our existing validator.

How to set up Schematron V2 Turbo in Codex

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

Use Spiral's MCP server

  1. Add the hosted server
    codex mcp add spiral --url https://api.spiral.ai/mcp
  2. Sign in

    Codex opens the browser: sign in to Spiral and click Authorize. No API key. codex mcp list then shows Spiral.

    codex mcp login spiral

Use the Spiral skill

  1. Send Codex 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.

    npx -y spiral-cli@latest init

Call the Spiral API

  1. Create an API key

    In the console, under API keys.

  2. Run Schematron V2 Turbo

    One request; the result comes back by polling or a webhook.

    curl https://api.spiral.ai/v1/runs \
      -H "Authorization: Bearer $SPIRAL_API_KEY" \
      -H "Content-Type: application/json" \
      -H "Idempotency-Key: $(uuidgen)" \
      -d '{
        "provider": "openrouter",
        "model": "inference-net/schematron-v2-turbo",
        "input": {
          "messages": [
            {
              "role": "system"
            }
          ]
        }
      }'

Then paste one of the prompts above: Codex finds Schematron V2 Turbo on Spiral, shows the price and runs it.

Schematron V2 Turbo providers and prices

Spiral runs Schematron V2 Turbo through OpenRouter. Open a card for today's price, with every fee included.

OpenRouterLLM
Inference.net: Schematron V2 Turbo
inference-net/schematron-v2-turbo

Settings you can ask for

Ask Codex for any of these in your prompt.

SettingAllowed valuesDefaultDescription
Seeda whole number
Min p0 to 1
Top ka whole number from 0
Top p0 to 1
Max tokens1 to 8192

The most output tokens to generate.

Temperature0 to 2
Presence penalty-2 to 2
Frequency penalty-2 to 2
Repetition penalty0 to 2

Every API field of Inference.net: Schematron V2 Turbo is on its tool page: open its card above.

What can you do with Schematron V2 Turbo in Codex?

Turn saved HTML into typed records for your project's ingestion pipeline. Have the coding agent define the extraction schema, clean the markup, and inspect the returned JSON.

Product catalog ingestion

Use Schematron V2 Turbo to extract product names, specifications, and category labels from product-page HTML into a consistent JSON structure.

Competitor availability monitoring

Extract product names and displayed availability from freshly fetched HTML. Compare the resulting records between runs to identify changes.

Company directory records

Turn company-page HTML into records containing organization names, website links, locations, and descriptions. Define each requested detail in the schema.

Research evidence extraction

Extract selected facts and supporting passages from retrieved HTML before another model writes an answer. Keep the source URL with each record.

Job listing aggregation

Extract job titles, employers, locations, and application links from listing HTML. Process large collections in smaller chunks to keep responses within the output limit.

Event calendar ingestion

Convert event-page HTML into structured titles, displayed dates, venues, and registration links. Preserve the original date text for review.

Article metadata collection

Extract headlines, bylines, publication dates, and article links from HTML for a searchable content index.

Questions about Schematron V2 Turbo in Codex

Can Codex use Schematron V2 Turbo?

Yes. Add Spiral to Codex as an MCP server, and Codex can find and run Schematron V2 Turbo like any built-in tool.

How much does Schematron V2 Turbo cost in Codex?

Up to $0.000801 per run at the default inputs. $0.039 per 1M input tokens, $0.195 per 1M output tokens. You pay from your Spiral balance as you go, with no subscription, and a failed run is not charged. Your Codex plan is separate.

Do I need an OpenRouter account to use Schematron V2 Turbo in Codex?

No. Spiral runs Schematron V2 Turbo on its own OpenRouter account. You sign in to Spiral once and never paste a key.

How long does it take to set up Schematron V2 Turbo in Codex?

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

Does Codex wait while Schematron V2 Turbo runs?

Schematron V2 Turbo runs in the background. Codex starts the run, checks on it and gives you the text, json when it's ready.

What is Schematron V2 best at?

Turning messy HTML into JSON with a defined structure. It fits product ingestion, page monitoring, and extracting selected facts for research.

How do I give Schematron V2 extraction instructions?

Put the directions in the JSON schema and its field descriptions. Send the HTML as the content to extract from; ordinary conversational instructions do not control the extraction.

How much HTML can Schematron V2 Turbo handle?

Input and output share a 128K-token context window, and the response can contain up to 8,192 tokens. Split large pages or long result lists into smaller extractions.

What temperature should I use?

Use temperature 0. Inference.net says Schematron was trained to perform best at that setting.

Should I clean the HTML first?

Yes. Remove scripts, styles, and boilerplate while preserving the content you need; Inference.net recommends lxml to match the training preprocessing.

Get started

Start using Schematron V2 Turbo in Codex

One Spiral account for every tool your agent needs. No subscription; you pay per run.

Paste this into Codex:

Set up Codex

set up https://spiral.ai/SKILL.md
Are you an AI agent? See setup options