DeepSeek's DeepSeek V4 Flash leads OpenRouter with 4.7T tokens (29 Jun 2026)

A leadership change on OpenRouter's Market Share ranking this week: google dethroned deepseek. Other rankings stayed mostly stable.

LLM Leaderboard

The big picture: which models AI builders are paying to run this week.

1. DeepSeek V4 Flash by DeepSeek: 4.7T tokens ↑4%
2. Hy3 preview by Tencent: 3.29T tokens ↑7%
3. Claude Opus 4.7 by Anthropic: 2.36T tokens ↑13%
4. DeepSeek V4 Pro by DeepSeek: 2.05T tokens ↑19%
5. Claude Opus 4.8 by Anthropic: 1.96T tokens ↑25%

Market Share

Vendor share of total OpenRouter traffic: who's winning the platform race.

1. google 798M tokens ↑28%
2. deepseek 596M tokens ↑20.9%
3. openai 450M tokens ↑15.8%
4. qwen 165M tokens ↑5.8%
5. anthropic 131M tokens ↑4.6%

Benchmark Leaders (AA Index)

Independent benchmark scores via Artificial Analysis. High AA Index = stronger reasoning, but cost matters too.

1. Claude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback) by Anthropic: 59.9 AA Index
2. Claude Opus 4.8 (Adaptive Reasoning, Max Effort) by Anthropic: 55.7 AA Index
3. GPT-5.5 (xhigh) by OpenAI: 54.8 AA Index
4. Claude Opus 4.7 (Adaptive Reasoning, Max Effort) by Anthropic: 53.5 AA Index
5. GPT-5.4 (xhigh) by OpenAI: 51.4 AA Index

Fastest Models

Throughput champs: pick these for latency-sensitive apps where speed beats raw quality.

1. gpt-oss-120b
2. Mercury 2
3. gpt-oss-safeguard-20b
4. Qwen3 32B
5. gpt-oss-20b

Top Models for English

Most-used models for English content this week: the multilingual leaders.

1. DeepSeek V4 Flash by DeepSeek: 748B tokens ↑11.8%
2. Hy3 preview by Tencent: 513B tokens ↑8.1%
3. DeepSeek V4 Pro by DeepSeek: 289B tokens ↑4.6%
4. Claude Opus 4.8 by Anthropic: 160B tokens ↑2.5%

Top Models for Python

Which models are getting picked for Python work right now.

1. DeepSeek V4 Flash by DeepSeek: 178B tokens ↑11%
2. Hy3 preview by Tencent: 149B tokens ↑9.2%
3. DeepSeek V4 Pro by DeepSeek: 72.2B tokens ↑4.5%
4. Claude Opus 4.8 by Anthropic: 32.1B tokens ↑2%

Top Models for short prompts (1K-10K tokens)

For short prompts (1K-10K tokens, the bulk of typical traffic), here's what builders chose.

1. DeepSeek V4 Flash by DeepSeek: 243M requests ↑17.3%
2. Gemini 2.5 Flash Lite by Google: 107M requests ↑7.6%
3. Gemini 2.5 Flash by Google: 88.1M requests ↑6.3%
4. Gemini 3 Flash Preview by Google: 62.6M requests ↑4.5%
5. Gemini 3.1 Flash Lite by Google: 60.6M requests ↑4.3%

Top Models for Tool Calls

If your stack uses tool calls / function calling, these models are getting the most invocations.

1. DeepSeek V4 Flash by DeepSeek: 51.6M tokens ↑10.3%
2. Hy3 preview by Tencent: 36.8M tokens ↑7.3%
3. DeepSeek V4 Pro by DeepSeek: 20.9M tokens ↑4.2%
4. Gemini 3 Flash Preview by Google: 19.1M tokens ↑3.8%
5. Claude Opus 4.7 by Anthropic: 15.1M tokens ↑3%

Top Image Models

Image-generation through OpenRouter: most-served models this week.

1. Gemini 2.5 Flash Lite by Google: 223M images ↑37.5%
2. Gemini 3 Flash Preview by Google: 35.3M images ↑5.9%
3. Gemini 2.5 Flash by Google: 31.4M images ↑5.3%
4. Claude Sonnet 4.6 by Anthropic: 29.6M images ↑5%
5. Qwen3.6 Plus by Qwen: 25.8M images ↑4.3%

Top Apps on OpenRouter

Useful as social proof when picking your stack: the largest public apps and agents that opt into OpenRouter usage tracking.

Most Popular

1. Hermes Agent (24.8T tokens): 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.
2. Kilo Code (6.75T tokens): 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.
3. OpenClaw (5.08T tokens): 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.
4. Claude Code (3.5T tokens): 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.

Trending

Fastest-growing apps on OpenRouter this week: early signals of which builder workflows are breaking out.

1. Hermes Agent (7.02T tokens) ↑18%
2. Claude Code (1.06T tokens) ↑31%
3. Cline (437B tokens) ↑82%
4. Pioneer (production) (441B tokens) ↑63%
5. Kilo Code (1.79T tokens) ↑9%
6. MavenBio (87.4B tokens) ↑761%
7. Cursor (114B tokens) ↑73%
8. timely-router

What this means for builders

The OpenRouter rankings are the cleanest signal of what AI builders pay to run, not what vendor marketing claims. This week the top of the leaderboard belongs to Google, and the fastest-growing app on the platform (Hermes Agent) is a useful tell about which builder workflows are quietly breaking out before the broader feed picks them up.