Token Counting
Estimate GPT-style token counts for long prompts and documents.
Context budgeting for long prompts
LLM Context Trimmer
Count tokens, reserve output space, and trim long text to fit an LLM context window.
Token BudgetTrim ModesCopy Prompt
Context trimmer
Trim long context to fit a model window, with response tokens reserved.
Prompt context
0 words · 0 tokens
Awaiting input—
Paste context to see whether it fits.
Usable 7,000
Budget8,000 tokens total
Prompt 0Reserved 1,000Headroom 7,000
Trimmed output
0 tokens · unchanged
Capabilities
What this tool handles
Use LLM Context Trimmer to check the input, tune the settings, and copy a clean output.
Smart Trimming
Keep the start, end, or both ends of a long context.
Prompt Ready
Reserve response tokens and copy a fitted prompt context.
Workflow
Run it in three steps
Fit long text into a practical token budget.
1
Paste Context
Add the transcript, article, code, or notes you want to send.
2
Set Budget
Choose context size, reserved output tokens, and trim mode.
3
Copy Trimmed Text
Use the fitted output in your next LLM prompt.
Good for
Common jobs this tool supports
Token EstimateCount GPT-style tokens
Trim ModesKeep start, end, or both
Prompt ReadyReserve room for responses
Local CountText stays in your browser
FAQ
Answers before you run it
Practical notes about output, privacy, and common settings.
It uses GPT-style tokenization as a practical estimate. Exact counts vary by model and provider.
Use next