Apps · 2026-10-02
The apps burning the tokens.
The AI apps routing the most LLM tokens through OpenRouter, and this week's breakouts. Data as of 2026-10-02.
Popular apps
| # | App | Category | Tokens (window) | Requests | Avg tok/req | 7d |
|---|---|---|---|---|---|---|
| 1 | Hermes Agent | coding | 49.70T | 505.21M | 98.4K | +7.5% |
| 2 | Claude Code | coding | 24.70T | 254.61M | 97.0K | -19.7% |
| 3 | Kilo Code | coding | 17.60T | 222.44M | 79.1K | +55.5% |
| 4 | Cline | coding | 16.34T | 137.08M | 119.2K | +26.6% |
| 5 | pi | coding | 10.17T | 97.48M | 104.3K | -4.4% |
| 6 | omp | coding | 8.21T | 48.02M | 170.9K | +2.9% |
| 7 | Codex | coding | 7.71T | 90.68M | 85.0K | +214.7% |
| 8 | OpenClaw | coding | 5.44T | 75.70M | 71.8K | -2.3% |
| 9 | DeepSeek Harness | — | 5.02T | 36.67M | 137.0K | +4.4% |
| 10 | Freebuff | — | 4.77T | 43.92M | 108.6K | +45.4% |
| 11 | Command Code | coding | 2.70T | 30.86M | 87.6K | +32.1% |
| 12 | OpenHands | coding | 2.09T | 21.88M | 95.4K | -65.9% |
| 13 | ZCode | coding | 1.67T | 9.12M | 183.4K | -19.3% |
| 14 | ISEKAI ZERO | entertainment | 1.66T | 34.72M | 47.8K | +28.4% |
| 15 | Hello Minds, powered by Ethoswarm | creative | 1.52T | 12.18M | 124.8K | -17.1% |
| 16 | HighLevel | — | 1.49T | 27.18M | 54.9K | +26.9% |
| 17 | Lemonade | coding | 1.34T | 30.39M | 43.9K | -1.3% |
| 18 | Framer | — | 1.32T | 12.00M | 109.7K | -40.8% |
| 19 | Strix | coding | 1.26T | 10.55M | 119.5K | +96.0% |
| 20 | Nous Research API | productivity | 1.22T | 44.48M | 27.4K | -7.0% |
| 21 | Descript | creative | 1.09T | 21.10M | 51.7K | +12.5% |
| 22 | Cursor | coding | 1.09T | 7.85M | 138.3K | +8.0% |
| 23 | draco-cascade-bench | — | 1.01T | 13.66M | 73.7K | +533.1% |
| 24 | CodeGPT | coding | 988.3B | 8.03M | 123.0K | -38.3% |
| 25 | Janitor AI | entertainment | 971.5B | 59.96M | 16.2K | -0.8% |
| 26 | Portkey AI | coding | 946.6B | 32.33M | 29.3K | -29.2% |
| 27 | Zed Editor | coding | 813.4B | 5.03M | 161.6K | +2.5% |
| 28 | Open WebUI | productivity | 771.4B | 237.29M | 3.3K | +29.4% |
| 29 | Proto Agent | coding | 669.9B | 2.41M | 278.1K | +22.9% |
| 30 | LangChain | — | 634.3B | 50.29M | 12.6K | +34.4% |
| 31 | HackerAI | — | 580.0B | 12.19M | 47.6K | +6.3% |
| 32 | Craft | creative | 534.8B | 28.63M | 18.7K | -2.1% |
| 33 | Zoo Code | — | 483.9B | 4.22M | 114.8K | -5.2% |
| 34 | Mira | creative | 449.4B | 29.02M | 15.5K | +41.4% |
| 35 | claude-mem | — | 410.4B | 10.33M | 39.7K | -0.9% |
| 36 | Letta | coding | 409.3B | 2.01M | 203.4K | -11.7% |
| 37 | GDevelop | coding | 404.1B | 7.21M | 56.1K | +4.9% |
| 38 | SillyTavern | entertainment | 365.2B | 18.71M | 19.5K | -3.1% |
| 39 | Ito | coding | 346.3B | 9.47M | 36.6K | -6.1% |
| 40 | Aura.bp.deepseekV4_1Flash | — | 333.2B | 3.28M | 101.5K | -1.2% |
| 41 | websim agent | — | 326.3B | 10.20M | 32.0K | +5.5% |
| 42 | ai-copilot-prod | — | 304.8B | 1.82M | 167.1K | -11.4% |
| 43 | MavenBio | — | 292.5B | 4.11M | 71.2K | +215.8% |
| 44 | Digen AI | coding | 283.2B | 18.07M | 15.7K | +6.4% |
| 45 | Sahasra | — | 251.4B | 6.48M | 38.8K | -13.7% |
| 46 | goose | coding | 224.1B | 2.90M | 77.4K | +130.3% |
| 47 | Roo Code | coding | 220.2B | 1.54M | 143.2K | +1.4% |
| 48 | gentity-fleet | — | 212.5B | 2.90M | 73.1K | -9.6% |
| 49 | Hello Haven | productivity | 212.1B | 7.43M | 28.6K | +8.3% |
| 50 | Halluna | productivity | 211.2B | 21.69M | 9.7K | -2.2% |
| 51 | Instaplay | entertainment | 197.6B | 971.7K | 203.4K | +464.4% |
| 52 | extra.email | productivity | 187.4B | 18.32M | 10.2K | +6.1% |
| 53 | Dayflow/90 CFNetwork/3860.700.1 Darwin/25.6.0 | — | 186.6B | 1.22M | 152.7K | -36.9% |
| 54 | pre.dev-agent | — | 184.4B | 3.29M | 56.1K | +375.8% |
| 55 | HammerAI | entertainment | 183.8B | 49.34M | 3.7K | +2.3% |
| 56 | HeyDitto | — | 182.9B | 26.94M | 6.8K | -36.0% |
| 57 | ale-evol | — | 177.6B | 3.11M | 57.1K | +650.9% |
| 58 | mimocode | coding | 175.9B | 989.1K | 177.9K | +35.5% |
| 59 | Pocket AI | — | 171.9B | 14.47M | 11.9K | +9.9% |
| 60 | Sophia's LoreBary | entertainment | 171.6B | 7.57M | 22.7K | -6.0% |
| 61 | jcode | — | 150.9B | 1.00M | 150.2K | +100.7% |
| 62 | Ouroboros | — | 150.4B | 593.2K | 253.5K | -79.6% |
| 63 | Qwen Code | coding | 147.6B | 1.14M | 129.9K | +11.8% |
| 64 | nanobot | productivity | 145.4B | 2.67M | 54.4K | -19.4% |
| 65 | Landingsite | — | 144.1B | 1.03M | 140.3K | -10.3% |
| 66 | AI Broker | — | 138.8B | 8.03M | 17.3K | -3.2% |
| 67 | Sentry MCP | — | 138.6B | 22.51M | 6.2K | -21.6% |
| 68 | ChainPatrol | — | 137.5B | 39.71M | 3.5K | -16.9% |
| 69 | LOVE | — | 130.2B | 5.33M | 24.4K | +18.9% |
| 70 | Crush | coding | 126.6B | 790.2K | 160.2K | -33.6% |
| 71 | Chatbox AI | productivity | 124.7B | 2.55M | 48.9K | +119.6% |
| 72 | Rosebud | — | 124.1B | 2.38M | 52.2K | -4.8% |
| 73 | Chub AI | entertainment | 121.1B | 9.02M | 13.4K | +0.9% |
| 74 | Trenzel | coding | 119.7B | 401.4K | 298.1K | +75.6% |
| 75 | Letaido | coding | 115.1B | 1.85M | 62.2K | +40.1% |
| 76 | Context7 | — | 114.9B | 44.13M | 2.6K | +2.1% |
| 77 | BobAI agents | — | 113.6B | 797.9K | 142.4K | +82.6% |
| 78 | v3-local-news-validation-api | — | 111.1B | 25.24M | 4.4K | +36.4% |
| 79 | Sekai | — | 110.4B | 27.46M | 4.0K | +104.0% |
| 80 | Cherry Studio | — | 109.7B | 1.39M | 79.0K | +6.2% |
| 81 | Tracer III RLVR Pipeline | — | 109.1B | 7.23M | 15.1K | +95.3% |
| 82 | ppq.ai | — | 109.0B | 2.36M | 46.1K | +15.9% |
| 83 | Open WebUI plugin for OpenRouter Responses API | productivity | 105.3B | 1.18M | 89.3K | +169.4% |
| 84 | Hoplite | — | 101.7B | 669.0K | 152.0K | -56.2% |
| 85 | unit-reason-graph | — | 101.1B | 10.15M | 10.0K | +38.6% |
| 86 | OrationAI | — | 99.8B | 2.83M | 35.2K | +35.4% |
| 87 | Miniapps.ai | — | 99.2B | 14.36M | 6.9K | +7.2% |
| 88 | ProVision.ai | — | 96.1B | 14.77M | 6.5K | -41.4% |
| 89 | OpenCode | — | 95.5B | 427.4K | 223.4K | +126.9% |
| 90 | LibreChat | productivity | 95.0B | 2.73M | 34.8K | +21.9% |
| 91 | KoalaBear | — | 94.4B | 19.00M | 5.0K | +115.6% |
| 92 | LangCC | — | 93.9B | 1.57M | 60.0K | +65.0% |
| 93 | Agent Zero | coding | 93.4B | 2.03M | 46.1K | -11.4% |
| 94 | Yeyu AI | — | 93.1B | 5.39M | 17.3K | +0.6% |
| 95 | Raylu | — | 91.4B | 24.41M | 3.7K | -8.0% |
| 96 | Eastman AI Gateway | — | 88.1B | 3.10M | 28.4K | +22.2% |
| 97 | Shelf Insights | — | 87.7B | 6.28M | 14.0K | -42.1% |
| 98 | Siege ORLoad | coding | 87.3B | 432.4K | 201.9K | +154.6% |
| 99 | My Dream Companion | — | 84.7B | 12.47M | 6.8K | +0.9% |
| 100 | OP.GG | productivity | 84.1B | 6.24M | 13.5K | +10.2% |
as of 2026-10-02
Heavy agents vs light chat
Every app placed by average tokens per request (x) and total window tokens (y), both log-scale.
Trenzel runs the heaviest average request at ~298k tokens/call.
How to read: Top-right = heavy agent workloads. Top-left = mass chatbot traffic. Bottom-right = niche high-context users. Bottom-left = light utilities.
What a session actually costs in Hermes Agent
Median cost of one coding session by session length. 30-day window, log scale.
Agent
GPT-6 Astra Pro costs 465× more per session than MiMo-V2.5 in this agent.
How to read: Each row is a model. Dots run from short sessions to long ones — the further right, the more a median session costs. The number on the right is the 10–49 turns median.
1 turn2–9 turns10–49 turns50+ turns
ModelMedian
- $8.94
- $4.02
- $3.52
- $3.05
- $2.80
- $2.06
- $1.60
- $1.37
- $1.30
- $1.08
