Free tools for anyone building with LLMs
Calculators, sortable references, and a public pricing API. Every tool reads the same dataset, so a price change shows up everywhere at once.
References
Bookmark-grade tables and matrices.
LLM pricing comparison
Sortable matrix of every major model — price, context, modalities, tokens-per-dollar.
OpenContext window comparison
What fits in 128K vs 1M vs 2M tokens — visualized.
OpenLLM speed benchmarks
Tokens-per-second and time-to-first-token across every major model.
OpenStructured outputs matrix
Which LLMs support JSON mode, strict schema, function calling, parallel tools.
OpenAPI rate limits cheatsheet
RPM and TPM by provider and tier — OpenAI, Anthropic, Gemini, Groq, more.
OpenDeveloper
Pull the data into your own tools.
Free downloads
Polished, no-signup downloads. The Claude Skills bundle drops straight into ~/.claude/skills/. The playbook is one long Markdown page with code per provider — read or print.
Download
Claude Skills that cut your LLM bill
Five drop-in Claude Code skills: cost auditor, prompt-cache optimizer, eval bootstrapper, model picker, live spend report. Stdlib Python, MIT, one zip.
Get the bundleRead
The $1k LLM bill rescue playbook
Ten specific things to do this week to halve an LLM bill — model swaps, prompt caching, max_tokens caps, batching, fallback chains. Real code per provider, ~10 min read.
Open the playbookGuides & comparisons
Longer-form content lives in three SEO hubs. None of it is gated, and we don’t harvest emails. Link to whatever’s useful.
How we keep them honest
Every tool reads from one canonical dataset (lib/llm-pricing-data.ts in the repo, /api/llm-prices.json in production). One file to update — every page and the public API reflects the change. No calculator can disagree with the reference table.
If you spot a wrong number, email [email protected] — we fix it once and it propagates everywhere.
Want a tool we don’t have?
Same address. If it’s a reasonable ask we’ll publish it.