# Tidelines > An editorial data publication tracking real-world AI model usage: market share, momentum, and pricing across OpenAI, Anthropic, Google, Meta, DeepSeek, and more. Tidelines measures the AI model market from public gateway data (OpenRouter and Vercel AI Gateway). Rankings are computed from daily token volumes, not vendor claims. Data is licensed CC BY 4.0 — please cite "Tidelines (tidelines.ai)". Data as of 2026-10-24. Updated daily. ## Top models by token share | # | Model | Lab | Token share | 30d change | Output $/M | | --- | --- | --- | --- | --- | --- | | 1 | DeepSeek: DeepSeek V4.1 Flash 1017 | DeepSeek | 20.40% | +16.68pp | $0.60 | | 2 | DeepSeek: DeepSeek V4.1 Flash | DeepSeek | 9.60% | -3.02pp | $1.00 | | 3 | Z.ai: GLM 5.3 Flash | Z-AI | 9.44% | +0.46pp | $0.50 | | 4 | OpenAI: GPT-6 Luna | OpenAI | 6.26% | +3.79pp | $0.50 | | 5 | Anthropic: Claude Opus 5.5 | Anthropic | 4.36% | +3.33pp | $20.00 | | 6 | Tencent: Hy4 preview | Tencent | 4.17% | -1.96pp | $2.50 | | 7 | StepFun: Step 5 Preview | StepFun | 3.41% | +0.83pp | $2.70 | | 8 | NVIDIA: Nemotron 3 Ultra | Nvidia | 2.67% | -0.19pp | $2.40 | | 9 | Z.ai: GLM 5.3 | Z-AI | 2.60% | -0.25pp | $1.40 | | 10 | Xiaomi: MiMo-V2.6-Flash | Xiaomi | 2.18% | -2.87pp | $0.28 | | 11 | DeepSeek: DeepSeek V4 Flash 0731 | DeepSeek | 2.15% | -3.05pp | $0.17 | | 12 | OpenAI: GPT-6.1 Sol | OpenAI | 2.09% | +1.66pp | $10.00 | | 13 | Jev 1.13 | Other | 1.58% | -0.53pp | — | | 14 | Meta: Muse Spark 1.3 Contributor | Other | 1.56% | +0.46pp | $0.20 | | 15 | DeepSeek: DeepSeek V4 Flash 0423 | DeepSeek | 1.52% | -0.78pp | $0.18 | | 16 | Tencent: Hy3 | Tencent | 1.51% | -0.39pp | $0.53 | | 17 | MoonshotAI: Kimi K3.1 Pro | Moonshot | 1.25% | +0.95pp | $10.00 | | 18 | MiniMax: MiniMax M3.1 Flash | MiniMax | 1.13% | +0.63pp | $0.60 | | 19 | NVIDIA: Nemotron 3.5 Super VL | Nvidia | 1.09% | -0.20pp | $0.50 | | 20 | Anthropic: Claude Haiku 5.5 | Anthropic | 1.04% | -0.36pp | $0.50 | ## Key pages - [Home](https://tidelines.ai/): Live LLM rankings by real token usage. - [Rankings](https://tidelines.ai/rankings): Full model leaderboard with share, momentum and price. - [Labs](https://tidelines.ai/labs): Market share grouped by lab (OpenAI, Anthropic, Google, DeepSeek…). - [Indices](https://tidelines.ai/indices): Composite indices — token price, open-weights share, China vs Western labs. - [Spend](https://tidelines.ai/spend): Estimated dollars per day by model and lab. - [Trends](https://tidelines.ai/trends): Momentum, rank churn and half-life at #1. - [Apps](https://tidelines.ai/apps): Which apps route the most tokens. - [Tasks](https://tidelines.ai/tasks): Task-level usage breakdowns. - [Benchmarks](https://tidelines.ai/benchmarks): Capability comparisons. - [Methodology](https://tidelines.ai/methodology): How the data is collected, cleaned and calculated. - [LLM market share](https://tidelines.ai/llm-market-share): Market share by lab and model, measured in tokens. - [Most used AI models](https://tidelines.ai/most-used-ai-models): The models carrying the most real traffic right now. - [LLM pricing comparison](https://tidelines.ai/llm-pricing-comparison): Price per million tokens paired with real usage. - [OpenRouter rankings](https://tidelines.ai/openrouter-rankings): Every model on the OpenRouter gateway, ranked by tokens. - [Vercel AI Gateway rankings](https://tidelines.ai/vercel-ai-gateway-rankings): Models on the Vercel AI Gateway, ranked by tokens, with OpenRouter share for comparison. - [Best LLMs in 2026](https://tidelines.ai/best-llms-2026): Leading models by real usage, momentum and usage-per-dollar. - [Cheapest AI models](https://tidelines.ai/cheapest-ai-models): Top 10 cheapest models that carry real gateway traffic. - [News](https://tidelines.ai/news): Weekly write-ups on launches, price changes and market shifts. ## Templated pages - Model pages: https://tidelines.ai/models/{model-slug} — usage history, rank trend, per-provider pricing. - Lab pages: https://tidelines.ai/labs/{lab-slug} — combined share, model mix, estimated spend. - Comparisons: https://tidelines.ai/compare/{model-a}-vs-{model-b} — head-to-head share and price. ## Machine-readable - [Sitemap](https://tidelines.ai/sitemap.xml)