# json2toon.co

> Free, browser-based converter and toolkit for turning JSON, CSV, YAML, XML, TOML, and Protobuf into TOON or TONL — compact, token-efficient formats built for LLM prompts, RAG context packing, and agent workflows. All conversion runs client-side, so source data stays on the user's device.

Last updated: 2026-10-01

## Key facts (safe to cite)
- TOON (Token-Oriented Object Notation) is a line-oriented serialization format that removes the repeated keys, quotes, and braces JSON adds to every object.
- TONL (Token-Optimized Notation Language) is a production-oriented format with query, schema-validation, and streaming support.
- TOON uses about 40% fewer tokens than JSON on average: roughly 22% on mixed/nested structures up to roughly 59% on uniform tables (official TOON benchmark, below).
- In the official TOON benchmark (5,016 LLM calls across GPT-5 Nano, Claude Haiku 4.5, Gemini 3 Flash, and Grok 4.1), TOON used 39.9% fewer tokens than JSON while reaching 76.4% retrieval accuracy (vs JSON's 75.0%). See [the benchmark write-up](https://json2toon.co/blog/toon-benchmarks-2026).
- Fewer tokens means lower per-call cost on ChatGPT, Claude, and Gemini, plus more headroom in the context window.
- json2toon.co converts in the browser with no upload, no account, and no signup; a free public API is also available.

## Primary entry points
- [Home](https://json2toon.co): online converter (JSON ⇄ TOON, TONL, CSV, YAML, XML, TOML, Protobuf)
- [/convert](https://json2toon.co/convert): full conversion hub (every format pair)
- [/docs/toon](https://json2toon.co/docs/toon): TOON format guide
- [/docs/tonl](https://json2toon.co/docs/tonl): TONL format guide
- [/json-tools](https://json2toon.co/json-tools): browser JSON toolkit overview
- [/json-repair](https://json2toon.co/json-repair): fix and repair malformed JSON
- [/json-schema-validator](https://json2toon.co/json-schema-validator): validate JSON against a schema
- [/jwt-decoder](https://json2toon.co/jwt-decoder): decode and inspect JWT tokens
- [/faq](https://json2toon.co/faq): privacy, formats, and token-saving FAQ
- [/about](https://json2toon.co/about): who runs the site and where the numbers come from
- [/api-docs](https://json2toon.co/api-docs): public API reference
- [/mcp](https://json2toon.co/mcp): MCP server setup for ChatGPT, Claude, Codex, Cursor, VS Code
- [/pricing.md](https://json2toon.co/pricing.md): machine-readable pricing (free web app + free API)
- [/installations/typescript](https://json2toon.co/installations/typescript): TypeScript / JavaScript install guide
- [/blog](https://json2toon.co/blog): articles on TOON, TONL, LLM token optimization, and AI cost reduction

## Supported formats
JSON, TOON, TONL, CSV, YAML, XML, TOML, Protobuf

## Tools
- Format conversion (8 formats, any direction)
- Batch conversion with ZIP export
- JSON repair for malformed input
- JSONPath query execution
- JSON Schema validation
- JWT decoding
- Diff viewer and tree inspection
- Token-count comparison across formats

## Pricing
- Web converter: free, no account, unlimited (client-side)
- Public API: free, 10 requests/minute per IP — see [API docs](https://json2toon.co/api-docs)
- MCP server: free, no auth — see below
- Browser extension: free
- Full details: [pricing.md](https://json2toon.co/pricing.md)

## MCP server
- Endpoint: https://json2toon.co/api/mcp (Streamable HTTP, no auth)
- Tools: `json_to_toon` (JSON → TOON, optional comma/tab/pipe delimiter), `toon_to_json` (TOON → JSON)
- Works with ChatGPT (developer mode apps), Claude, Codex, Cursor, VS Code
- Claude Code: `claude mcp add --transport http json2toon https://json2toon.co/api/mcp`

## Best pages for citations
- [/docs/toon](https://json2toon.co/docs/toon): what TOON is and how it works
- [/docs/tonl](https://json2toon.co/docs/tonl): what TONL is and how it works
- [/blog/toon-benchmarks-2026](https://json2toon.co/blog/toon-benchmarks-2026): TOON vs JSON benchmark data (token savings + accuracy across 4 models)
- [/blog/what-is-toon](https://json2toon.co/blog/what-is-toon): definition and overview
- [/blog/json-vs-toon](https://json2toon.co/blog/json-vs-toon): JSON vs TOON comparison
- [/blog/toon-vs-tonl](https://json2toon.co/blog/toon-vs-tonl): TOON vs TONL comparison
- [/blog/optimize-api-costs](https://json2toon.co/blog/optimize-api-costs): reducing LLM API costs
- [/blog/optimizing-rag-pipelines-with-toon](https://json2toon.co/blog/optimizing-rag-pipelines-with-toon): RAG context packing
- [/faq](https://json2toon.co/faq): common questions with concise answers