OpenRouter app leaderboard: who burns the most on LLM inference
OpenRouter publishes per-app token counts for every AI tool routing through it. We read the public app listings, join each model to live OpenRouter pricing, and rank the top 30 apps by estimated 30-day spend. Cost assumes a 72/28 input/output split — a reasonable approximation for tool-heavy agentic workloads.
33.48T
Total tokens (30d)
$74.62M
Monthly cost (top 30)
601.8M
Total requests
32
Apps tracked
Three things the data says
- OpenClaw is the biggest line item by far at $29.96M/month — 40.2% of all spend across the top 30 apps. It moves 20.24T tokens at a blended $1.76/M.
- HammerAI runs the leanest at $0.03/M blended — roughly 186.6× cheaper per token than VidMuse at $6.39/M. Same job, different model routing.
- Token volume ≠ dollars. Claude Code moves 3.51T (more than Hermes Agent's 1.96T) but spends $14.90M vs Hermes's $3.18M. Model mix dominates.
Ranked by monthly cost
Each app's top-20 model mix, priced against OpenRouter's live catalog. A few models (stealth/alpha lanes, unlisted internal routes) carry no price and are excluded from the dollar total but still count toward the token total.
| # | App | Categories | Tokens | Requests | Monthly cost | Blended $/M |
|---|---|---|---|---|---|---|
| 1 | 20.24T | 342.1M | $29.96M | $1.76 | ||
| 2 | 3.51T | 50.7M | $14.90M | $4.50 | ||
| 3 | 6.29T | 103.9M | $8.75M | $1.44 | ||
| 4 | 274.4B | 5.6M | $6.67M | $5.79 | ||
| 5 | 1.96T | 33.5M | $3.18M | $1.80 | ||
| 6 | 157.6B | 2.2M | $2.78M | $1.85 | ||
| 7 | programming-app | 111.6B | 1.5M | $1.65M | $3.72 | |
| 8 | roleplay | 26.4B | 568K | $695K | $6.39 | |
| 9 | 51.7B | 3.2M | $669K | $3.70 | ||
| 10 | 74.4B | 3.1M | $608K | $2.07 | ||
| 11 | roleplay | 35.1B | 2.1M | $605K | $4.22 | |
| 12 | 199.1B | 12.8M | $603K | $0.75 | ||
| 13 | 4.5B | 227K | $569K | $3.96 | ||
| 14 | 175.9B | 14.6M | $545K | $0.75 | ||
| 15 | 43.4B | 989K | $481K | $5.17 | ||
| 16 | 47.5B | 1.4M | $324K | $2.45 | ||
| 17 | 3.6B | 32K | $284K | $6.02 | ||
| 18 | 14.8B | 1.7M | $206K | $3.17 | ||
| 19 | 29.2B | 2.4M | $203K | $1.90 | ||
| 20 | 28.1B | 1.7M | $201K | $3.91 | ||
| 21 | 6.2B | 184K | $152K | $6.19 | ||
| 22 | 13.2B | 239K | $126K | $2.14 | ||
| 23 | personal-agent | 23.4B | 492K | $89K | $0.82 | |
| 24 | 20.4B | 385K | $85K | $1.38 | ||
| 25 | roleplaygame | 19.4B | 923K | $82K | $1.73 | |
| 26 | 46.0B | 765K | $82K | $1.73 | ||
| 27 | cli-agent | 4.4B | 208K | $79K | $4.88 | |
| 28 | 4.2B | 159K | $22K | $1.08 | ||
| 29 | personal-agent | 5.8B | 87K | $14K | $0.72 | |
| 30 | 49.1B | 13.5M | $7K | $0.03 |
Ranked by raw token volume
Full tracked app list including tail entries without model breakdowns. Volume-only ranking — no cost join.
| # | App | Global rank | Tokens | Requests |
|---|---|---|---|---|
| 1 | OpenClaw | #1 | 20.24T | 342.1M |
| 2 | Kilo Code | #2 | 6.29T | 103.9M |
| 3 | Claude Code | #3 | 3.51T | 50.7M |
| 4 | Hermes Agent | #7 | 1.96T | 33.5M |
| 5 | Descript | #8 | 274.4B | 5.6M |
| 6 | ISEKAI ZERO | #10 | 199.1B | 12.8M |
| 7 | Janitor AI | #11 | 175.9B | 14.6M |
| 8 | Cline | #12 | 157.6B | 2.2M |
| 9 | Roo Code | #13 | 111.6B | 1.5M |
| 10 | Lemonade | #15 | 74.4B | 3.1M |
| 11 | SillyTavern | #17 | 51.7B | 3.2M |
| 12 | HammerAI | #18 | 49.1B | 13.5M |
| 13 | OpenHands | #19 | 47.5B | 1.4M |
| 14 | Nous Research API | #20 | 46.0B | 765K |
| 15 | Ito | #21 | 43.4B | 989K |
| 16 | Agent Zero | #24 | 35.1B | 2.1M |
| 17 | Chub AI | #25 | 29.2B | 2.4M |
| 18 | LangChain | #28 | 28.1B | 1.7M |
| 19 | VidMuse | #30 | 26.4B | 568K |
| 20 | Gobii | #33 | 23.4B | 492K |
| 21 | Nexu Link | #35 | 20.4B | 385K |
| 22 | Goose | #40 | 19.4B | 923K |
| 23 | Portkey AI | #46 | 14.8B | 1.7M |
| 24 | Zed Editor | #50 | 13.2B | 239K |
| 25 | novelcrafter | #66 | 6.2B | 184K |
| 26 | Qwen Code | #75 | 5.8B | 87K |
| 27 | BLACKBOXAI | #91 | 4.5B | 227K |
| 28 | LibreChat | #93 | 4.4B | 208K |
| 29 | ZeroClaw | #96 | 4.2B | 159K |
| 30 | Codex | #105 | 3.6B | 32K |
| 31 | openbench | #115 | 3.3B | 561K |
| 32 | Aider | #140 | 2.6B | 82K |
Methodology
- Data source:
openrouter.ai/appsand each app's detail page, fetched via the React Server Components stream (RSC: 1header). Captured 2026-04-14. - Pricing: public
openrouter.ai/api/v1/modelscatalog. Permaslugs (e.g.anthropic/claude-4.6-opus-20260205) are matched back to base catalog IDs after stripping date and preview suffixes. - Blended $/M = price_in × 0.72 + price_out × 0.28. Agentic workloads are context-heavy because every tool round replays the conversation. Flipping to 50/50 inflates dollar totals by ~10–15%.
- Top-20 models per app: OpenRouter surfaces the top 20 models by tokens for each app. Apps may use more models beyond that slice at long-tail volume, so reported spend is a lower bound.
- Unpriced models: stealth/alpha lanes (e.g.
openrouter/hunter-alpha) and unlisted providers are dropped from the dollar total. Free tiers are priced at $0. - Not a quality ranking. This page measures compute spend, not output quality. For task performance see our agentic benchmarks.
- Deep dives: inverted model view · category rollups · Hermes Agent case study. Every app in the ranked tables links to its own per-app breakdown page.
App missing from the leaderboard?
Know an app routing serious volume through OpenRouter that isn't here? Send the slug — we'll include it in the next refresh.