# The Token Company

> LLM input compression middleware. Remove low-signal tokens from prompts to cut costs, improve accuracy, and fit more context.

## About

The Token Company builds compression models that remove low-signal tokens from LLM prompts before they reach the model. The compression is semantic and deterministic - it processes tokens based on context and intent, not simple truncation. Backed by Y Combinator, with investors from Supercell, Wolt, Hugging Face, Silo AI, and SVA.

Website: https://thetokencompany.com

## Installation

- Python: `pip install the-token-company`
- TypeScript: `npm install the-token-company`

## Models

### bear-2
Latest and recommended compression model. Most accurate compression quality.

### bear-1.2
Previous generation. Faster compression, lower latency per request.

## How It Works

1. Send text to The Token Company compression API
2. Receive compressed text back
3. Pass the compressed text to any LLM (GPT, Claude, Gemini, etc.)

That's the entire integration - one API call before your LLM call. The SDK provides `with_compression()` (Python) and `withCompression()` (TypeScript) wrappers for OpenAI and Anthropic that handle this automatically.

## Pricing

Free tier: up to 1B processed tokens per month, bear-2 & bear-1.2 models, API access.
Enterprise: custom pricing for unlimited tokens, custom fine-tuned models, on-premise hosting, dedicated support, SLA guarantees, and more.

Details: https://thetokencompany.com/pricing

## Use Cases

- **LLM Entertainment & Gaming**: Longer memories, richer worlds, same budget
- **Meeting Transcription Analysis**: Distill hours of calls into signal-dense context
- **Web Scraping**: Strip boilerplate from crawled pages before ingest
- **Document Analysis**: Fit more PDFs and reports into one context window

## Case Study: Pax Historia

Pax Historia processes 193 billion tokens per month on OpenRouter, making them one of the biggest token consumers globally. They ran a 268K-vote blind model arena with bear-1.1 compression.

Results:
- +4.9% improvement in Sonnet 4.5 score
- +15% improvement in Gemini 3 Flash score
- +5% purchase amount lift in A/B tests

Full case study: https://thetokencompany.com/blog/pax-historia

## Benchmarks

### FinanceBench
Accuracy evaluation on real-world financial documents, not synthetic benchmarks.

Details: https://thetokencompany.com/benchmarks/financebench

## Documentation

- [Introduction](https://thetokencompany.com/docs): What is TTC, why compress, quick example
- [Quickstart](https://thetokencompany.com/docs/quickstart): Install SDK, get API key, first compression
- [About Compression](https://thetokencompany.com/docs/compression): Models, aggressiveness levels, how it works
- [OpenAI](https://thetokencompany.com/docs/openai): Automatic compression for OpenAI API calls
- [Vercel AI SDK](https://thetokencompany.com/docs/vercel-ai): Middleware compression for any AI SDK provider
- [Anthropic](https://thetokencompany.com/docs/anthropic): Automatic compression for Claude API calls
- [OpenRouter](https://thetokencompany.com/docs/openrouter): Compression for OpenRouter and OpenAI-compatible providers
- [Protect Text](https://thetokencompany.com/docs/protect-text): Exclude content from compression
- [Data Retention](https://thetokencompany.com/docs/data-retention): Zero data retention policy
- [Full docs as plain text](https://thetokencompany.com/docs/llms.txt): All documentation in a single text file for LLMs

## Key Links

- [Home](https://thetokencompany.com)
- [Pricing](https://thetokencompany.com/pricing)
- [Blog](https://thetokencompany.com/blog)
- [Benchmarks](https://thetokencompany.com/benchmarks)
- [Contact](https://thetokencompany.com/contact)
- [Careers](https://thetokencompany.com/careers)
- [Privacy Policy](https://thetokencompany.com/privacy)

## Blog Posts

- [Pax Historia Case Study](https://thetokencompany.com/blog/pax-historia): One of the biggest token consumers globally improved quality by removing context bloat
- [bear-1.1 Release](https://thetokencompany.com/blog/bear-1-1): Improved accuracy preservation and faster compression speeds
- [bear-1 Launch](https://thetokencompany.com/blog/bear-1): First LLM input compression model