SuperAPI AI Directory
Partner with us中文AI newsFavorites

AI coding

2,798 results · Page 78 of 78

Tokenade

AI coding

Tokenade is a token optimization tool for AI coding agents, designed to reduce token consumption during operation by agents such as Claude Code, Cursor, Codex, and Copilot. It compresses noisy output such as builds, tests, logs, and MCP tool results, keeping only the key information before the model reads it; it also provides capabilities such as an MCP proxy, agent-specific web search, local code semantic search, prompt streamlining, context budget management, and cache reuse. Tokenade emphasizes local-first operation: it does not upload code, prompts, or file paths, uploads only aggregated statistical data, and by default redacts secret information. It supports one-command installation, and free accounts can save up to 10 million tokens per month, Pro 100 million per month, and Enterprise unlimited.

Read more

TokenCalculator.dev

AI coding

TokenCalculator.dev is a browser-based LLM token counting and API cost estimation tool. Users can paste text prompts, or attach PDF, DOCX, code, data files, as well as PNG, JPEG, WebP, GIF images, and view in real time the token count, context window fit, and estimated input cost for models across providers. The tool supports 43 current models from 9 providers, including OpenAI, Anthropic, Google, DeepSeek, Qwen, Moonshot Kimi, xAI Grok, Z.ai GLM, and Meta. All text extraction, image dimension reading, and token calculation are done locally in the browser, without uploading user content, with a focus on privacy protection. OpenAI text uses exact local o200k_base BPE tokenization; other providers use a deterministic provider-calibrated UTF-8 token projection, and the results indicate the method and accuracy. Pricing data comes from Models.dev with an offline fallback.

Read more