GrindLab .glrange Format: Multi-Action Ranges, Done Right

Most poker range files only encode one action per combo. You export a range from GTO Wizard or Hand2Note, you get a CSV like AhKs: 0.6, KhQs: 0.4 — and that's the frequency for one action (raise, or call, or 3bet). The rest of the strategy is implicit.

GrindLab's .glrange format is different: it stores the full multi-action breakdown (raise + call + 3bet + fold all in one file) so you can roundtrip a complex GTO range without losing information. This article explains the format, when to use it, and how it compares to the standard text exports of other tools.

The problem: text exports flatten multi-action ranges

Open any modern solver and look at the strategy view. A hand like AKs might be:

• 60% raise
• 30% call
• 10% 3bet

The UI shows this as a split cell with three colors. The data exists internally. But when you click "Export range", what you get is a text file with one frequency per combo — typically the action you have selected. To capture the full mix, you'd have to export three separate files.

This is fine for sharing a single decision (e.g. "what's your CO open range?"). It falls apart when you want to share a complete strategy or back up your work. Information leaks at every export.

The .glrange format: lossless multi-action storage

.glrange is a JSON-based bundle format used by GrindLab to import and export ranges. Each range stores:

The actionFreqs field is the key. It mirrors how solvers store their internal data — frequency per action per hand. Roundtripping a range through .glrange preserves the full mix.

A bundle can also contain multiple ranges plus the folder hierarchy that organizes them. So you can ship an entire "MTT 100bb" study pack in a single file.

Multi-action vs single-action: a side-by-side

To be clear: every tool above supports multi-action ranges internally. Their UIs render the splits properly. The limitation is in their text export format — it's a lossy interchange layer designed for compatibility, not fidelity.

How GrindLab handles imports from other tools

GrindLab's import flow accepts both formats and auto-detects which you're pasting. The single-action limitation of the source format becomes a per-import constraint:

1. From the proprietary .glrange (full multi-action)

Drop a .glrange file in the Import modal → '.glrange' tab. The bundle is parsed, all multi-action metadata is preserved, ranges land in your library exactly as the sender built them.

2. From a solver text export (single action)

Drop a .txt, .csv, or .zip in the 'From solver' tab. GrindLab auto-detects the format:

• Combo-level (AhKs:0.5, KdQs:1) — Hand2Note / GTO Wizard style
• Hand-level (99+, A2s+, [50]TT[/50]) — Flopzilla / Equilab style

Each file becomes one range. You pick the action it represents (Raise / Call / 3Bet / etc.) and the imported range goes into your library tagged with that action. To recover a full multi-action mix, import each action separately and combine them in the Editor.

3. ZIP support

Drop a .zip containing multiple range files. GrindLab expands it inline (one slot per file), filters out OS metadata (__MACOSX/, ._*, .DS_Store), and lets you configure each range's name + action in a single batch.

The case for the proprietary format

.glrange is proprietary, which has obvious tradeoffs:

Pros

• Preserves multi-action data losslessly
• Carries notes, folder structure, custom action palettes
• Compact (entire study packs in <1 MB)
• Versioned format with validation — won't silently corrupt
• Single file = easy to email/Slack/Discord

Cons

• Only GrindLab can read it directly
• For sending a range to a non-GrindLab user, you have to export each action as a separate text file

The proprietary format is the right tool when you control both ends of the pipe (your own backup, sharing with another GrindLab user, coaching deliveries). The single-action text formats are the right tool when you need to feed a range into a different ecosystem.

Use cases by workflow

"I'm building my study library from scratch"

→ Use .glrange exclusively for backups. Build ranges in the Editor with full multi-action assignments. Export periodically as a backup. Never lose data on browser cache wipe.

"I'm coaching students"

→ Build the range pack in GrindLab → export .glrange → students import it. They get the exact same ranges with your annotations, action assignments, folder organization. No interpretation lost.

"I'm migrating from Flopzilla/Equilab"

→ Use the Import → 'From solver' tab. Paste your existing range strings (99+, A2s+, KJo+, etc.). They become single-action ranges in GrindLab. Re-tag actions if needed. → Read the GrindLab vs Flopzilla comparison for migration tips.

"I'm using GTO Wizard solutions in my prep"

→ Export the relevant action from GTO Wizard → drop the file in GrindLab. Repeat for each action you need. Combine in the Editor for a multi-action master range. → See GrindLab vs GTO Wizard for the full integration story.

"I want to share a range pack with my study group"

→ Bundle the ranges into a folder → Export → send the .glrange file. Everyone imports it the same way. No format negotiation needed.

Inside the .glrange file: what it actually contains

A .glrange file is just JSON. You can open it in any text editor. Roughly:

{
  "format": "grindlab.range.bundle",
  "version": 1,
  "exportedAt": "2026-05-03T12:00:00Z",
  "exportedBy": { "app": "GrindLab", "appVersion": "1.0" },
  "ranges": [
    {
      "name": "CO Open 100bb",
      "hands": ["AA", "AKs", "AKo", "..."],
      "weights": { "AKo": 0.5 },
      "actionFreqs": { "AKs": { "raise": 0.6, "call": 0.4 } },
      "actionDefs": [
        { "id": "raise", "name": "Raise", "color": "#3FB950" },
        { "id": "call",  "name": "Call",  "color": "#58A6FF" }
      ],
      "notes": "MP open vs LJ flat call: raise weak suited Ax."
    }
  ],
  "folders": [{ "id": "...", "name": "MTT 100bb", "parentId": null }]
}

The format is documented and validated server-side. Future versions will be backward-compatible — bundles you export today will still import in a year.

Limits to be aware of

• Hold'em only. The format is designed for 2-card holdings on a 13×13 grid. PLO (4-card) is on the roadmap but not yet supported.
• Bundle size cap: ~10 MB to keep the API fast. A typical study pack is well under 1 MB.
• No live state: .glrange only stores ranges, not equity calculations or in-progress hand analysis. Those live in their own data model.

What to use when

Summary

The .glrange format exists because no standard text export captures multi-action frequencies. By making the format proprietary and JSON-based, GrindLab can preserve the full strategy mix per range — including notes, folder structure, and custom action palettes — in a single file.

For exchange with other ecosystems, the import flow accepts the standard combo-level and hand-level text formats used by Hand2Note, GTO Wizard, Flopzilla and Equilab. Each imported range is single-action by design (the source format's limitation), but you can layer multiple imports in the Editor to reconstruct a multi-action master range.

The proprietary format is the right tool when fidelity matters. The text formats are the right tool when interoperability matters. Both have their place.

→ Read the complete GrindLab guide for an overview of the rest of the toolkit, or check how to build poker ranges for the strategy side.