Rankings

What the world is actually running.

Real aggregate usage across the open model ecosystem — token volume, requests, and market share, refreshed regularly. Pick a model the market trusts — or pick the underdog it hasn't noticed yet.

40 models · usage as of 2026-05-22

Top models

#ModelProviderModalityTokensRequestsShare
1Claude Opus 4.7
anthropic/claude-opus-4.7
Anthropictext2.75T65.2M22.9%
2Claude Sonnet 4.6
anthropic/claude-sonnet-4.6
Anthropictext2.13T63.1M17.7%
3Gemini 3 Flash
google/gemini-3-flash
Googletext1.05T127.5M8.7%
4Gemini 2.5 Flash Lite
google/gemini-2.5-flash-lite
Googletext649.2B259.9M5.4%
5Kimi-K2.6
moonshot/kimi-k2.6
Moonshottext641.3B19.9M5.3%
6Gpt-Oss-120b
openai/gpt-oss-120b
OpenAItext566.6B112.1M4.7%
7Gemini 2.5 Flash
google/gemini-2.5-flash
Googletext565.1B149.4M4.7%
8Gemini 3.1 Flash Lite
google/gemini-3.1-flash-lite
Googletext548.1B109.4M4.6%
9GPT-5.5
openai/gpt-5.5
OpenAItext498.9B9.7M4.2%
10Claude Opus 4.6
anthropic/claude-opus-4.6
Anthropictext483.1B9.6M4%
11GPT-5.4
openai/gpt-5.4
OpenAItext335.8B14.5M2.8%
12Gemini 3.1 Pro
google/gemini-3.1-pro
Googletext298.5B17.8M2.5%
13Claude Haiku 4.5
anthropic/claude-haiku-4.5
Anthropictext243.8B28.7M2%
14Gemma-4-26b-A4b-IT
google/gemma-4-26b-a4b-it
Googletext201.9B57.6M1.7%
15GPT-5.4 Mini
openai/gpt-5.4-mini
OpenAItext158.4B20.2M1.3%
16Claude Sonnet 4.5
anthropic/claude-sonnet-4.5
Anthropictext124.1B8.7M1%
17GPT-5.4 Nano
openai/gpt-5.4-nano
OpenAItext115.1B17.1M1%
18GPT-5
openai/gpt-5
OpenAItext91.3B7.4M0.8%
19Gpt-Oss-20b
openai/gpt-oss-20b
OpenAItext86.6B27.6M0.7%
20Grok 4.3
xai/grok-4.3
xAItext80.6B12.2M0.7%
21GPT-4.1 Mini
openai/gpt-4.1-mini
OpenAItext78.5B27.1M0.7%
22Gemini 2.5 Pro
google/gemini-2.5-pro
Googletext71.6B6.7M0.6%
23Qwen 3.5 397B A17B
alibaba/qwen3.5-397b-a17b
Alibabatext67.0B7.5M0.6%
24GPT-4.1
openai/gpt-4.1
OpenAItext48.7B11.3M0.4%
25Claude Sonnet 4
anthropic/claude-sonnet-4
Anthropictext47.9B3.7M0.4%
26Claude Opus 4.8
anthropic/claude-opus-4.8
Anthropictext39.9B644.2K0.3%
27Llama-4-Scout-17b-16e-Instruct
meta/llama-4-scout-17b-16e-instruct
Metatext11.7B6.8M0.1%
28Gemma-3-12b-IT
google/gemma-3-12b-it
Googletext7.5B11.5M0.1%
29Bge-M3
baai/bge-m3
BAAIembedding6.1B7.1M0.1%
30GPT-5.5 Pro
openai/gpt-5.5-pro
OpenAItext5.9B292.5K0%
31Seedream 4.5
bytedance/seedream-4.5
ByteDanceimage2.9B147.3K0%
32O4-Mini
openai/o4-mini
OpenAItext2.0B645.3K0%
33Qwen 3 Max
alibaba/qwen3-max
Alibabatext1.5B251.9K0%
34Llama-3.2-3b-Instruct
meta/llama-3.2-3b-instruct
Metatext1.4B4.7M0%
35GPT-5.4 Pro
openai/gpt-5.4-pro
OpenAItext1.3B164.2K0%
36Grok Imagine Image Quality
xai/grok-imagine-image-quality
xAIimage703.6M130.3K0%
37Llama-3-8b-Instruct
meta/llama-3-8b-instruct
Metatext692.6M880.9K0%
38FLUX.2 [Pro] Preview
black-forest-labs/flux-2-pro-preview
Black Forest Labsimage676.9M53.4K0%
39Bge-Base-EN-V1.5
baai/bge-base-en-v1.5
BAAIembedding652.4M1.5M0%
40Llama-3.2-1b-Instruct
meta/llama-3.2-1b-instruct
Metatext450.2M270.0K0%

By provider

ProviderModelsTokens
Anthropic75.82T
Google93.39T
OpenAI131.99T
Moonshot1641.3B
xAI281.3B
Alibaba268.5B
Meta414.2B
BAAI36.8B
ByteDance12.9B
Black Forest Labs41.4B
Ibm-granite1400.0M
Deepseek1290.1M
How this is built

Real usage, not guesswork.

These rankings reflect real aggregate usage across the open model ecosystem, refreshed regularly — real market signal, not invented numbers. As first-party traffic on AIgateway grows, it's blended into the same board — opt your app in from Dashboard → Settings.