Market share · 2026-10-08
LLM market share
Share of real API traffic, measured in tokens across public AI gateways. Not survey data, not vendor claims — the tokens that actually moved yesterday.
Leading lab
28.8%
DeepSeek
Leading model
22.64%
DeepSeek: DeepSeek V4.1 Flash
Open-weight share
63.8%
of tracked tokens
Tokens / day
23T
50 models
The state of the LLM market in October 2026
DeepSeek holds 28.8% of tracked tokens across 5 models, +10.86pp over 30 days. Z-AI follows at 11.1%, then OpenAI at 9.5%. Open-weight models together account for 63.8% of traffic — the single clearest signal of how much of the market is price-driven rather than brand-driven.
Token share and dollar share are different markets. The tracked models represent about $51.1M per day at list prices, and the labs at the top of the token table are rarely the labs at the top of the spend table. The spend view separates the two.
These figures are free to cite under CC BY 4.0. Attribute to “Tidelines (tidelines.ai),2026-10-08”.
Market share by lab
Share of gateway tokens on the latest archived day, all models combined.
| # | Lab | Share % | 30d Δ | Models | Top model |
|---|---|---|---|---|---|
| 1 | DeepSeek | 28.77 | +10.86pp | 5 | DeepSeek: DeepSeek V4.1 Flash |
| 2 | Z-AI | 11.07 | -2.55pp | 3 | Z.ai: GLM 5.3 Flash |
| 3 | OpenAI | 9.46 | -5.05pp | 7 | OpenAI: GPT-6 Luna |
| 4 | Xiaomi | 8.23 | -0.71pp | 3 | Xiaomi: MiMo-V2.6-Flash |
| 5 | Tencent | 7.20 | -13.69pp | 2 | Tencent: Hy4 preview |
| 6 | Anthropic | 6.88 | +1.69pp | 5 | Anthropic: Claude Opus 5.5 |
| 7 | Other | 6.87 | +1.17pp | 9 | Jev 1.13 |
| 8 | Nvidia | 4.88 | +1.42pp | 3 | NVIDIA: Nemotron 3 Ultra |
| 9 | 3.55 | -2.67pp | 8 | Google: Gemini 3.8 Flash | |
| 10 | StepFun | 2.59 | +0.00pp | 1 | StepFun: Step 5 Preview |
| 11 | Moonshot | 1.34 | -0.01pp | 1 | MoonshotAI: Kimi K3 |
| 12 | MiniMax | 0.82 | -0.36pp | 1 | MiniMax: MiniMax M3 |
| 13 | xAI | 0.70 | +0.41pp | 1 | SpaceXAI: Grok 4.7 |
| 14 | Alibaba | 0.26 | -0.07pp | 1 | Qwen: Qwen3.8 Max (0902) |
Market share by model
Top 20 individual models by share of gateway tokens.
| # | Model | Lab | Share % | 30d Δ |
|---|---|---|---|---|
| 1 | DeepSeek: DeepSeek V4.1 Flash | DeepSeek | 22.64 | +19.11pp |
| 2 | Z.ai: GLM 5.3 Flash | Z-AI | 7.48 | -2.01pp |
| 3 | Xiaomi: MiMo-V2.6-Flash | Xiaomi | 7.16 | +6.44pp |
| 4 | Tencent: Hy4 preview | Tencent | 5.29 | -12.58pp |
| 5 | OpenAI: GPT-6 Luna | OpenAI | 5.21 | +4.85pp |
| 6 | NVIDIA: Nemotron 3 Ultra | Nvidia | 3.95 | +1.45pp |
| 7 | Anthropic: Claude Opus 5.5 | Anthropic | 3.35 | +2.37pp |
| 8 | DeepSeek: DeepSeek V4 Flash 0731 | DeepSeek | 2.73 | -6.31pp |
| 9 | StepFun: Step 5 Preview | StepFun | 2.59 | +0.00pp |
| 10 | Z.ai: GLM 5.3 | Z-AI | 2.36 | +0.08pp |
| 11 | Jev 1.13 | Other | 2.02 | +1.58pp |
| 12 | DeepSeek: DeepSeek V4 Flash 0423 | DeepSeek | 1.94 | -1.60pp |
| 13 | Tencent: Hy3 | Tencent | 1.91 | -1.11pp |
| 14 | Meta: Muse Spark 1.3 Contributor | Other | 1.54 | -0.11pp |
| 15 | Anthropic: Claude Haiku 5.5 | Anthropic | 1.40 | +0.00pp |
| 16 | MoonshotAI: Kimi K3 | Moonshot | 1.34 | -0.01pp |
| 17 | OpenAI: GPT-6.1 Sol | OpenAI | 1.28 | +0.85pp |
| 18 | Google: Gemini 3.8 Flash | 1.26 | -1.71pp | |
| 19 | OpenAI: GPT-5.6 Luna | OpenAI | 1.24 | -9.39pp |
| 20 | Z.ai: GLM 5.2 | Z-AI | 1.23 | -0.62pp |
LLM market share FAQ
- Which AI lab has the largest LLM market share?
- DeepSeek leads with 28.8% of tracked gateway tokens as of 2026-10-08, ahead of Z-AI at 11.1%.
- How is LLM market share measured here?
- By tokens, not revenue or headlines. Tidelines archives daily token volumes published by public AI gateways (OpenRouter and Vercel AI Gateway) and computes each model's and lab's share of that total. It is a measure of developer API traffic, not of consumer chatbot users.
- How many models and labs are tracked?
- 50 models across 14 labs on the latest archived day (2026-10-08), representing roughly 23T tokens per day.
- Can I cite these market share numbers?
- Yes. The data is published under CC BY 4.0 — cite "Tidelines (tidelines.ai)" with the date of the snapshot you used.