Apps · 2026-07-15

The apps burning the tokens.

The AI apps routing the most LLM tokens through OpenRouter, and this week's breakouts. Data as of 2026-07-15.

Popular apps

Tidelines.ai
#AppCategoryTokens (window)RequestsAvg tok/req7d
1Hermes Agentcoding27.49T357.07M77.0K
2Kilo Codecoding7.54T116.74M64.6K
3Claude Codecoding5.18T60.30M86.0K
4OpenClawcoding4.87T88.41M55.1K
5Clinecoding2.16T21.10M102.1K
6Descriptcreative2.06T26.68M77.2K
7picoding1.99T37.15M53.4K
8Janitor AIentertainment956.3B69.61M13.7K
9Pioneer (production)931.0B16.30M57.1K
10ISEKAI ZEROentertainment906.2B21.45M42.2K
11Lemonadecoding890.8B25.83M34.5K
12Pieces.app: Ambient Artificial Memoryproductivity561.8B62.29M9.0K
13Miraproductivity441.1B25.91M17.0K
14Cursorcoding424.8B3.81M111.4K
15MavenBio414.4B6.18M67.1K
16LangChain346.0B58.50M5.9K
17GDevelopcoding345.6B9.65M35.8K
18Endpoint Proxy310.7B7.17M43.4K
19Oh-My-Picoding308.7B2.86M108.0K
20HackerAI302.6B7.17M42.2K
21Codexcoding299.0B4.04M74.0K
22Zoo Code290.9B3.60M80.8K
23SillyTavernentertainment281.2B16.53M17.0K
24Zed Editorcoding272.5B2.80M97.3K
25Command Codecoding249.5B8.54M29.2K
26Nous Research APIproductivity238.0B14.56M16.3K
27HammerAIentertainment225.7B61.30M3.7K
28Portkey AIcoding220.6B17.02M13.0K
29Peezy Gatewaycoding217.1B3.69M58.9K
30extra.emailproductivity204.6B29.12M7.0K
31OpenHandscoding198.9B4.23M47.0K
32Roo Codecoding190.7B1.80M106.2K
33OpenManus Low-Cost Agent186.8B12.22M15.3K
34Open WebUIproductivity168.2B10.16M16.6K
35Craftcreative167.3B26.19M6.4K
36Hello Minds, powered by Ethoswarmcreative163.6B4.10M39.9K
37OpenCX162.3B15.56M10.4K
38Sahasra148.3B2.74M54.2K
39Codebuffcoding147.4B16.93M8.7K
40Sophia's LoreBaryentertainment145.4B7.23M20.1K
41shapes incentertainment141.0B37.33M3.8K
42Z Code137.7B1.21M114.1K
43Chub AIentertainment123.8B9.31M13.3K
44Hallunaproductivity115.2B9.46M12.2K
45Landingsite114.8B1.24M92.7K
46SquadStack AI111.6B17.86M6.2K
47llmv:snappy_summary101.8B42.16M2.4K
48ppq.ai101.2B2.90M35.0K
49OpenSquillacoding97.7B2.95M33.1K
50LOVE95.8B2.89M33.2K
51Letaidocoding93.8B1.36M68.9K
52goosecoding90.7B2.43M37.3K
53timely-router88.2B9.09M9.7K
54Pocket AI83.7B8.73M9.6K
55GitLawbcoding81.0B1.31M61.7K
56claude-mem79.6B3.54M22.5K
57Agent Zerocoding79.6B2.52M31.5K
58JobLeads LLMproductivity75.7B18.63M4.1K
59My Dream Companion74.9B8.45M8.9K
60Miniapps.ai74.9B12.60M5.9K
61Qwen Codecoding73.5B1.18M62.2K
62Ethoswarmproductivity69.7B1.79M38.8K
63SkyrimNetentertainment67.1B16.18M4.1K
64Factory AI Droid66.2B527.1K125.5K
65LangCC65.9B1.95M33.8K
66OP.GGproductivity61.9B4.36M14.2K
67Poolsidecoding59.9B1.25M47.8K
68Itocoding58.7B1.67M35.2K
69One API58.5B16.25M3.6K
70BuddyPro AIproductivity56.8B4.66M12.2K
71SiriusAI Multilingual Data Pipelinecreative55.1B4.08M13.5K
72Skyvern54.8B5.87M9.3K
73MS.ai CoS Chat54.4B1.64M33.3K
74Internal LLM Proxy53.9B1.35M40.0K
75Orchid53.7B7.70M7.0K
76Crushcoding50.3B1.13M44.6K
77Dify.AIproductivity50.1B21.04M2.4K
78OrationAI48.5B1.29M37.7K
79Hermes Web UIcoding48.1B374.8K128.4K
80Continue43.4B1.00M43.2K
81ChainPatrol42.3B28.20M1.5K
82ThetaWave42.0B12.02M3.5K
83payready-document-extraction41.5B16.29M2.5K
84KoalaBear41.3B18.10M2.3K
85LibreChatproductivity40.6B1.91M21.2K
86mimocodecoding40.6B400.6K101.3K
87Plato Computer Use Agent38.7B700.0K55.3K
88TeleClawcoding38.5B1.29M29.8K
89Covena AI38.2B1.97M19.4K
90Bitsec36.8B3.22M11.4K
91nanobotproductivity36.5B1.41M25.9K
92unit-reason-graph36.0B5.09M7.1K
93SkyrimNetentertainment33.8B3.36M10.1K
94Yeyu AI33.5B2.27M14.8K
95Yumina33.4B996.2K33.5K
96Cursor IDE Model Test32.6B341.9K95.2K
97Screener AI30.4B1.08M28.1K
98FVChat29.6B11.42M2.6K
99AI Tool Calling Benchmark29.6B359.7K82.2K
100Manifest29.5B652.5K45.2K

as of 2026-07-15

Heavy agents vs light chat

Every app placed by average tokens per request (x) and total window tokens (y), both log-scale.

Tidelines.ai

Hermes Web UI runs the heaviest average request at ~128k 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.

1k10k100k1M10B100B1T10T100Theavy agentsmass chatbotsniche power userslight utilitiesAvg tokens / request (log)Total window tokens (log)Hermes Agent — 77k tok/req · 27T totalKilo Code — 65k tok/req · 8T totalClaude Code — 86k tok/req · 5T totalOpenClaw — 55k tok/req · 5T totalCline — 102k tok/req · 2T totalDescript — 77k tok/req · 2T totalpi — 53k tok/req · 2T totalJanitor AI — 14k tok/req · 956B totalPioneer (production) — 57k tok/req · 931B totalISEKAI ZERO — 42k tok/req · 906B totalLemonade — 34k tok/req · 891B totalPieces.app: Ambient Artificial Memory — 9k tok/req · 562B totalMira — 17k tok/req · 441B totalCursor — 111k tok/req · 425B totalMavenBio — 67k tok/req · 414B totalLangChain — 6k tok/req · 346B totalGDevelop — 36k tok/req · 346B totalEndpoint Proxy — 43k tok/req · 311B totalOh-My-Pi — 108k tok/req · 309B totalHackerAI — 42k tok/req · 303B totalCodex — 74k tok/req · 299B totalZoo Code — 81k tok/req · 291B totalSillyTavern — 17k tok/req · 281B totalZed Editor — 97k tok/req · 273B totalCommand Code — 29k tok/req · 249B totalNous Research API — 16k tok/req · 238B totalHammerAI — 4k tok/req · 226B totalPortkey AI — 13k tok/req · 221B totalPeezy Gateway — 59k tok/req · 217B totalextra.email — 7k tok/req · 205B totalOpenHands — 47k tok/req · 199B totalRoo Code — 106k tok/req · 191B totalOpenManus Low-Cost Agent — 15k tok/req · 187B totalOpen WebUI — 17k tok/req · 168B totalCraft — 6k tok/req · 167B totalHello Minds, powered by Ethoswarm — 40k tok/req · 164B totalOpenCX — 10k tok/req · 162B totalSahasra — 54k tok/req · 148B totalCodebuff — 9k tok/req · 147B totalSophia's LoreBary — 20k tok/req · 145B totalshapes inc — 4k tok/req · 141B totalZ Code — 114k tok/req · 138B totalChub AI — 13k tok/req · 124B totalHalluna — 12k tok/req · 115B totalLandingsite — 93k tok/req · 115B totalSquadStack AI — 6k tok/req · 112B totalllmv:snappy_summary — 2k tok/req · 102B totalppq.ai — 35k tok/req · 101B totalOpenSquilla — 33k tok/req · 98B totalLOVE — 33k tok/req · 96B totalLetaido — 69k tok/req · 94B totalgoose — 37k tok/req · 91B totaltimely-router — 10k tok/req · 88B totalPocket AI — 10k tok/req · 84B totalGitLawb — 62k tok/req · 81B totalclaude-mem — 22k tok/req · 80B totalAgent Zero — 32k tok/req · 80B totalJobLeads LLM — 4k tok/req · 76B totalMy Dream Companion — 9k tok/req · 75B totalMiniapps.ai — 6k tok/req · 75B totalQwen Code — 62k tok/req · 73B totalEthoswarm — 39k tok/req · 70B totalSkyrimNet — 4k tok/req · 67B totalFactory AI Droid — 126k tok/req · 66B totalLangCC — 34k tok/req · 66B totalOP.GG — 14k tok/req · 62B totalPoolside — 48k tok/req · 60B totalIto — 35k tok/req · 59B totalOne API — 4k tok/req · 59B totalBuddyPro AI — 12k tok/req · 57B totalSiriusAI Multilingual Data Pipeline — 14k tok/req · 55B totalSkyvern — 9k tok/req · 55B totalMS.ai CoS Chat — 33k tok/req · 54B totalInternal LLM Proxy — 40k tok/req · 54B totalOrchid — 7k tok/req · 54B totalCrush — 45k tok/req · 50B totalDify.AI — 2k tok/req · 50B totalOrationAI — 38k tok/req · 48B totalHermes Web UI — 128k tok/req · 48B totalContinue — 43k tok/req · 43B totalChainPatrol — 1k tok/req · 42B totalThetaWave — 3k tok/req · 42B totalpayready-document-extraction — 3k tok/req · 41B totalKoalaBear — 2k tok/req · 41B totalLibreChat — 21k tok/req · 41B totalmimocode — 101k tok/req · 41B totalPlato Computer Use Agent — 55k tok/req · 39B totalTeleClaw — 30k tok/req · 39B totalCovena AI — 19k tok/req · 38B totalBitsec — 11k tok/req · 37B totalnanobot — 26k tok/req · 36B totalunit-reason-graph — 7k tok/req · 36B totalSkyrimNet — 10k tok/req · 34B totalYeyu AI — 15k tok/req · 34B totalYumina — 34k tok/req · 33B totalCursor IDE Model Test — 95k tok/req · 33B totalScreener AI — 28k tok/req · 30B totalFVChat — 3k tok/req · 30B totalAI Tool Calling Benchmark — 82k tok/req · 30B totalManifest — 45k tok/req · 29B totalHermes Web UI
Data: OpenRouter apps as of 2026-07-15

Token counts come from each provider's own tokenizer. Cross-app comparisons are approximate — treat them as ordinal, not absolute.

Source: OpenRouter (openrouter.ai/apps), as of 2026-07-15.