OmniGlyph

diegosouzapw/OmniGlyph is an open-source project by diegosouzapw: Cut your Claude bill 59–70% by rendering bulky LLM context as dense PNG pages — 100% read accuracy, exact per-provider billing math (Anthropic/OpenAI/Gemini), fail-closed gates, measured benchmarks. CLI + proxy + Cloudflare Workers. Docs in 42 languages. Part of the OmniRoute family.. Indexed here so it can be found — not resold.

It is free. Get it from the upstream repository: https://github.com/diegosouzapw/OmniGlyph

From the project's own README (excerpt, reproduced for discovery under its MIT license):

🖼️ OmniGlyph — Context as Image

Cut your Claude bill by 59–70% by rendering bulky context as dense PNG pages — the same content, in a fraction of the tokens.

Models bill text per token, but bill an image by its dimensions — not by how much text is inside it.

Part of the OmniRoute family · 🌐 All languages

🇺🇸 🇸🇦 🇦🇿 🇧🇬 🇧🇩 🇨🇿 🇩🇰 🇩🇪 🇪🇸 🇮🇷 🇫🇮 🇫🇷 🇮🇳 🇮🇱 🇮🇳 🇭🇺 🇮🇩 🇮🇹 🇯🇵 🇰🇷 🇮🇳 🇲🇾 🇳🇱 🇳🇴 🇵🇭 🇵🇱 🇵🇹 🇧🇷 🇷🇴 🇷🇺 🇸🇰 🇸🇪 🇰🇪 🇮🇳 🇮🇳 🇹🇭 🇹🇷 🇺🇦 🇵🇰 🇻🇳 🇨🇳 🇹🇼

Eight years of context growth, in characters. Every text line tops out near ~4M chars (a 1M-token window at ~4 chars/token); the orange point is the same Fable 5 1M window read through OmniGlyph images — ~18M chars at the measured 18.16 chars/vision-token, 4.54× the text ceiling. The density and multiplier are measured from a live render at generation time, not hand-typed: regenerate with npx tsx scripts/gen-context-chart.ts (source).

---

📊 The numbers — measured, not estimated

metric result receipt --- --- --- End-to-end bill reduction 59–70% production trace, 13,709 requests Tokens per converted block 10× fewer (28,080 chars: 14,040 → 1,460 tokens) billing sweep Billing-formula accuracy residual zero across 22 counttokens probes, 2 models × 2 tiers benchmarks/billing-sweep/results/ Exact-read accuracy, production config 30/30 (100%) on Claude Fable 5 density frontier Silent confabulations in ~300 read probes 0 — every miss abstains as ILEGIVEL benchmarks/density-frontier/results/

Model scorecard (can it read dense renders? n=30 per arm, deterministic scoring):

model reading verdict --- --- --- Claude…