Ranked by real token throughput on the OpenRouter marketplace - not benchmarks. Useful for seeing which models people actually run.
Tokens
167.26T
Requests
8.5B
Models
548
Lead
DeepSeek V4.1 Flash
DeepSeek
Token mix this window
98% prompt · 2% completion
Share of token volume this window
15.4%
↓31%
9.8%
-
6.8%
↓12%
6.2%
-
5.1%
↓25%
5.1%
↑9%
4.3%
↑26%
3.6%
↓14%
3.1%
↑18%
2.7%
↓40%
Showing 20 of 548
| # | Model | Type | Tokens | Share | Requests | Tools | Price / M | Context | Δ wk |
|---|---|---|---|---|---|---|---|---|---|
| 1 | DeepSeek V4.1 Flash DeepSeek · OpenRouter | Vision | 25.7T 25.2T in · 490.6B out | 15.4% | 643.5M | - | $0.30 / $1.20 | 1.0M | ↓31% |
| 2 | DeepSeek V4.1 Flash 1017 DeepSeek · OpenRouter | Vision | 16.5T 16.2T in · 236.3B out | 9.8% | 284.0M | - | $0.30 / $1.20 | 1.0M | - |
| 3 | GLM 5.3 Flash Z.ai · OpenRouter | Vision | 11.3T 11.0T in · 360.9B out | 6.8% | 664.1M | - | $0.02 / $0.12 | 1.0M | ↓12% |
| 4 | MiniMax M3.1 Flash (free) MiniMax · free · OpenRouter | Vision | 10.3T 10.2T in · 111.6B out | 6.2% | 84.2M | - | Free | 1M | - |
| 5 | MiMo-V2.6-Flash Xiaomi · OpenRouter | Audio | 8.57T 8.47T in · 99.8B out | 5.1% | 112.9M | - | $0.14 / $0.28 | 1.1M | ↓25% |
| 6 | Hy4 preview Tencent · OpenRouter | Language | 8.55T 8.46T in · 88.6B out | 5.1% | 81.5M | - | $0.83 / $2.50 | 1.0M | ↑9% |
| 7 | Claude Opus 5.5 Anthropic · OpenRouter | Vision | 7.14T 7.07T in · 67.6B out | 4.3% | 54.8M | - | $4.00 / $20 | 1M | ↑26% |
| 8 | GPT-6 Luna OpenAI · OpenRouter | Vision | 6.08T 5.86T in · 221.0B out | 3.6% | 431.3M | - | $0.10 / $0.50 | 1.1M | ↓14% |
| 9 | GLM 5.3 Z.ai · OpenRouter | Language | 5.21T 5.13T in · 83.4B out | 3.1% | 69.1M | - | $0.08 / $1.40 | 1.0M | ↑18% |
| 10 | Nemotron 3 Ultra (free) NVIDIA · free · OpenRouter | Language | 4.46T 4.43T in · 25.6B out | 2.7% | 31.5M | - | Free | 1M | ↓40% |
| 11 | DeepSeek V4 Flash 0731 DeepSeek · OpenRouter | Language | 4.29T 4.10T in · 193.6B out | 2.6% | 360.5M | - | $0.44 / $1.32 | 1.0M | ↓9% |
| 12 | Vision | 3.18T 3.14T in · 41.2B out | 1.9% | 54.1M | - | $0.10 / $0.20 | 1.0M | ↑6% | |
| 13 | Hy3 Tencent · OpenRouter | Language | 3.09T 3.04T in · 48.3B out | 1.8% | 39.1M | - | $0.13 / $0.53 | 262K | ↓4% |
| 14 | jev-1.13-20260917 Typesafe · OpenRouter | Language | 2.93T 2.71T in · 216.2B out | 1.8% | 947.9M | - | Free | - | ↓12% |
| 15 | DeepSeek V4 Flash 0423 DeepSeek · OpenRouter | Language | 2.77T 2.59T in · 183.3B out | 1.7% | 329.8M | - | $0.09 / $0.18 | 1.0M | ↓12% |
| 16 | Claude Sonnet 5.5 Anthropic · OpenRouter | Vision | 2.53T 2.49T in · 36.4B out | 1.5% | 42.7M | - | $2.00 / $10 | 1M | ↓7% |
| 17 | Step 5 Preview StepFun · OpenRouter | Vision | 2.32T 2.31T in · 12.4B out | 1.4% | 12.1M | - | $1.00 / $2.70 | 1.0M | ↓88% |
| 18 | Claude Haiku 5.5 Anthropic · OpenRouter | Vision | 2.09T 1.97T in · 118.5B out | 1.2% | 135.0M | - | $0.10 / $0.50 | 1M | ↑7% |
| 19 | GPT-6.1 Sol OpenAI · OpenRouter | Vision | 1.98T 1.95T in · 25.7B out | 1.2% | 46.3M | - | $2.00 / $10 | 1.1M | ↓14% |
| 20 | Kimi K3 Moonshot · OpenRouter | Vision | 1.72T 1.69T in · 30.8B out | 1.0% | 29.4M | - | $0.24 / $12 | 1.0M | ↓35% |
Weekly request share by model author
Models that actually invoke tools
Images processed through OpenRouter
Public apps by token volume this week
Claude Code
Claude Code is Anthropic's agentic coding tool that reads your entire codebase, plans and executes changes across files, runs tests, and iterates on failures, all from natural language prompts.
Hermes Agent
Hermes Agent is an open-source, self-improving AI agent by Nous Research that runs persistently with memory across sessions, and builds reusable skills from experience. It comes with 40+ built-in tools, including web search, browser automation, and vision, plus scheduled automations and subagents.
Kilo Code
Kilo Code is an open-source AI coding agent that works across VS Code, JetBrains, and CLI to help developers ship code faster with agentic workflows.
pi
There are many coding agents, but this one is yours.
Cline
Cline is an open-source AI coding agent that lives inside your IDE, autonomously exploring your codebase, editing files, running terminal commands, and using browser automation.
DeepSeek Harness
Freebuff
Powerful coding models, funded by ads
omp
OpenClaw
OpenClaw is an open-source AI agent that connects to your messaging apps and takes real actions on your behalf, from running commands and browsing the web to managing files and sending emails.
Codex
A coding agent that helps you build and ship with AI
Proto Agent
Proto is the personal agent for your business.
Strix
Source: OpenRouter rankings, as of 2026-10-21. Licensed under CC BY 4.0. Price is prompt / completion per million tokens. First-party app and direct vendor traffic are not included.