About Aitools Test
A framework-free, zero-server AI text analysis toolkit — 5-metric AI detection, tiktoken-based token counting with output estimation, and user-defined LLM cost comparison. Built by 345tool • June 15, 2026
Pure Client-Side AI Text Analysis Without Server Overhead
Aitools Test is built by the 345tool developer collective — an independent team engineering lightweight, stateless browser utilities. The core philosophy is simple: AI content detection, token estimation, and API cost projection should execute locally in your browser, not on someone else's server. Every algorithm — perplexity scoring, burstiness analysis, tiktoken approximation, and cost calculation — runs inside your browser's JavaScript sandbox with zero network round-trips during analysis.
Core Engineering Features
Three analysis dimensions operate inside a single text input — no tab switching, no API keys, no server processing. Paste once, see everything.
AI Text Detector: 5-Dimensional Statistical Analysis
The detector evaluates text across five orthogonal dimensions: perplexity variance (word-pair frequency distributions), burstiness patterns (sentence length variation), n-gram repetition density, type-token vocabulary ratio, and structural uniformity. Each metric feeds a weighted composite score from 0% (human-written) to 100% (AI-generated), with per-metric bar visualizations so you can see which features drive the result. Best performance on English prose above 300 characters.
Token Counter: Input Counting + Output Estimation
The tokenizer approximates the tiktoken cl100k_base algorithm with CJK-aware splitting — roughly 0.75 tokens per English word and 1.5 per CJK character. Beyond basic input counting, the engine detects questions and task-style prompts in your text, then estimates likely output tokens based on complexity (simple, moderate, complex, or heavy). This gives you a projected total — input + output — which is essential for context window planning and cost forecasting. Estimates fall within 5–10% of official tiktoken output.
Cost Estimator: User-Defined Model Pricing
Instead of shipping with stale, hardcoded price tables, the cost estimator lets you define your own per-1M-token rates for any LLM — add GPT-4.1, Claude 4 Sonnet, Gemini 2.5 Pro, DeepSeek V3, or custom fine-tuned endpoints via the "+ Add Model" button. Costs are computed as (input + estimated output tokens) / 1M × your rate, with results sorted cheapest-first. Only models with a price entered appear in the comparison view, so you can focus on what matters for your workload.
Technical Architecture & Privacy Guarantees
Every analysis algorithm — detection, tokenization, and cost projection — executes inside your browser's local JavaScript runtime. No text content ever traverses the network during analysis.
Zero-Server Data Footprint
When you paste or type text, all computations run on your device. Nothing is transmitted to any remote server, stored in a database, or logged by third-party analytics. Disconnect your internet after page load and every feature continues to work offline. This architecture aligns with the 345tool privacy mandate of absolute client-side data sovereignty.
Offline-Capable Design
With zero external API dependencies, Aitools Test is fully functional offline once page assets load. This makes it suitable for analyzing sensitive documents, proprietary business communications, and unpublished manuscripts in environments where network access is restricted or data confidentiality is paramount. Analysis results are transient — they vanish when you close or refresh the tab, leaving no persistent records.
Contact
We are the 345tool Team
345tool is an independent developer collective engineering elite, pure client-side, and privacy-first web utilities to replace bloated internet tools.