Compare AI providers across score, stability, momentum, and market share. Rankings derived from composite scores of 300 tracked models, updated hourly.
| # | Provider | Models | Avg Score | Best Model | Top Rank | 24h Avg | 7d Avg | Stability% | Free |
|---|---|---|---|---|---|---|---|---|---|
| 1 | Anthropic | 15 | 83.7 | Claude Fable 5(96.6) | #1 | +0.3 | +0.1 | 100% | 0 |
| 2 | xAI | 4 | 74.0 | Grok 4.20(88.3) | #17 | 0 | 0 | 100% | 0 |
| 3 | Xiaomi | 2 | 74.0 | MiMo-V2.5-Pro(75.6) | #65 | +1.0 | 0 | 100% | 0 |
| 4 | DeepSeek | 11 | 71.5 | DeepSeek V4 Pro(86.2) | #32 | +0.7 | +0.1 | 100% | 0 |
| 5 | 22 | 69.8 | Gemini 3.1 Pro Preview Custom Tools(91.7) | #7 | +0 | -0.1 | 100% | 4 | |
| 6 | OpenAI | 58 | 69.8 | GPT-5.5(92.2) | #6 | +0.5 | +0.2 | 100% | 2 |
| 7 | Zhipu AI | 12 | 68.3 | GLM 5.2(78.1) | #56 | -11.3 | +9.7 | 92% | 0 |
| 8 | MiniMax | 8 | 67.7 | MiniMax M2.5(77.6) | #57 | +0.9 | 0 | 100% | 0 |
| 9 | Cursor | 2 | 65.3 | Composer 2(65.3) | #121 | +1.0 | 0 | 100% | 0 |
| 10 | Inception | 1 | 60.5 | Mercury 2(60.5) | #138 | +1.0 | 0 | 100% | 0 |
| 11 | Moonshot AI | 6 | 57.2 | Kimi K2.6(75.2) | #67 | +1.0 | +0.7 | 100% | 0 |
| 12 | Microsoft | 2 | 56.1 | Phi 4(59.9) | #142 | +1.0 | +0.5 | 100% | 0 |
| 13 | Meta | 8 | 55.4 | Llama 4 Maverick(67.4) | #108 | +0.8 | +0.3 | 100% | 2 |
| 14 | Allen AI | 1 | 54.5 | Olmo 3 32B Think(54.5) | #151 | +1.0 | +1.0 | 100% | 0 |
| 15 | Tencent | 2 | 54.0 | Hy3 preview(67.9) | #106 | +0.5 | 0 | 100% | 0 |
| 16 | StepFun | 2 | 53.3 | Step 3.5 Flash(66.6) | #112 | +0.5 | 0 | 100% | 0 |
| 17 | Alibaba | 48 | 52.3 | Qwen3.5 397B A17B(79.1) | #47 | +0.4 | 0 | 100% | 2 |
| 18 | IBM | 2 | 47.5 | Granite 4.1 8B(55.0) | #149 | +0.5 | +0.5 | 100% | 0 |
| 19 | Mistral AI | 18 | 47.1 | Mistral Medium 3.5(71.2) | #85 | -0.3 | -0.5 | 94% | 0 |
| 20 | Cohere | 4 | 46.8 | Command A(50.4) | #161 | +1.0 | +1.0 | 100% | 1 |
| 21 | arcee-ai | 4 | 45.5 | Trinity Large Thinking(62.7) | #135 | 0 | -1.3 | 100% | 0 |
| 22 | Amazon | 5 | 44.0 | Nova 2 Lite(60.1) | #140 | +0.2 | 0 | 100% | 0 |
| 23 | NVIDIA | 11 | 41.8 | Llama 3.3 Nemotron Super 49B V1.5(60.1) | #141 | -0.3 | -0.4 | 100% | 7 |
| 24 | sakana | 1 | 40.0 | Fugu Ultra(40.0) | #170 | +147.0 | +147.0 | 0% | 0 |
| 25 | ~anthropic | 4 | 40.0 | Claude Fable Latest(40.0) | #172 | 0 | 0 | 100% | 0 |
| 26 | perceptron | 1 | 40.0 | Perceptron Mk1(40.0) | #180 | 0 | 0 | 100% | 0 |
| 27 | inclusionai | 3 | 40.0 | Ring-2.6-1T(40.0) | #181 | 0 | 0 | 100% | 0 |
| 28 | poolside | 4 | 40.0 | Laguna XS.2 (free)(40.0) | #184 | 0 | 0 | 100% | 2 |
| 29 | ~openai | 2 | 40.0 | OpenAI GPT Mini Latest(40.0) | #189 | 0 | 0 | 100% | 0 |
| 30 | 2 | 40.0 | Google Gemini Pro Latest(40.0) | #190 | 0 | 0 | 100% | 0 | |
| 31 | ~moonshotai | 1 | 40.0 | MoonshotAI Kimi Latest(40.0) | #191 | 0 | 0 | 100% | 0 |
| 32 | Kuaishou | 1 | 40.0 | KAT-Coder-Pro V2(40.0) | #205 | -1.0 | -1.0 | 100% | 0 |
| 33 | rekaai | 2 | 40.0 | Reka Edge(40.0) | #206 | -0.5 | -0.5 | 100% | 0 |
| 34 | ByteDance | 5 | 40.0 | Seed-2.0-Lite(40.0) | #210 | -0.8 | -0.8 | 100% | 0 |
| 35 | Liquid AI | 3 | 40.0 | LFM2-24B-A2B(40.0) | #212 | -1.0 | -1.0 | 100% | 2 |
| 36 | aion-labs | 3 | 40.0 | Aion-2.0(40.0) | #213 | -0.3 | -0.3 | 100% | 0 |
| 37 | Upstage | 1 | 40.0 | Solar Pro 3(40.0) | #216 | -1.0 | -1.0 | 100% | 0 |
| 38 | Writer | 1 | 40.0 | Palmyra X5(40.0) | #217 | -1.0 | -1.0 | 100% | 0 |
| 39 | deepcogito | 1 | 40.0 | Cogito v2.1 671B(40.0) | #230 | 0 | 0 | 100% | 0 |
| 40 | Perplexity | 5 | 40.0 | Sonar Pro Search(40.0) | #232 | 0 | 0 | 100% | 0 |
| 41 | AI21 Labs | 1 | 40.0 | Jamba Large 1.7(40.0) | #248 | 0 | 0 | 100% | 0 |
| 42 | Baidu | 1 | 40.0 | ERNIE 4.5 VL 424B A47B (40.0) | #256 | 0 | 0 | 100% | 0 |
| 43 | Windsurf | 1 | 40.0 | SWE-1.5(40.0) | #287 | +1.0 | +1.0 | 100% | 0 |
| 44 | TII | 6 | 40.0 | Falcon-H1-Arabic 34B Instruct(40.0) | #290 | +1.0 | +1.0 | 100% | 7 |
| 45 | HUMAIN | 3 | 39.5 | ALLaM 2 7B Instruct(40.0) | #288 | +0.7 | +0.7 | 100% | 2 |
Providers with at least 2 models, ranked by percentage of models in a "stable" state. Higher stability means more consistent performance over time.
Distribution of tracked models across providers. Shows each provider's share of the total 300 models in the leaderboard.
| Provider | Models | Share | % |
|---|---|---|---|
| OpenAI | 58 | 19.3% | |
| Alibaba | 48 | 16.0% | |
| 22 | 7.3% | ||
| Mistral AI | 18 | 6.0% | |
| Anthropic | 15 | 5.0% | |
| Zhipu AI | 12 | 4.0% | |
| DeepSeek | 11 | 3.7% | |
| NVIDIA | 11 | 3.7% | |
| MiniMax | 8 | 2.7% | |
| Meta | 8 | 2.7% | |
| Moonshot AI | 6 | 2.0% | |
| TII | 6 | 2.0% | |
| Amazon | 5 | 1.7% | |
| ByteDance | 5 | 1.7% | |
| Perplexity | 5 | 1.7% | |
| xAI | 4 | 1.3% | |
| Cohere | 4 | 1.3% | |
| arcee-ai | 4 | 1.3% | |
| ~anthropic | 4 | 1.3% | |
| poolside | 4 | 1.3% | |
| inclusionai | 3 | 1.0% | |
| Liquid AI | 3 | 1.0% | |
| aion-labs | 3 | 1.0% | |
| HUMAIN | 3 | 1.0% | |
| Xiaomi | 2 | 0.7% | |
| Cursor | 2 | 0.7% | |
| Microsoft | 2 | 0.7% | |
| Tencent | 2 | 0.7% | |
| StepFun | 2 | 0.7% | |
| IBM | 2 | 0.7% | |
| ~openai | 2 | 0.7% | |
| 2 | 0.7% | ||
| rekaai | 2 | 0.7% | |
| Inception | 1 | 0.3% | |
| Allen AI | 1 | 0.3% | |
| sakana | 1 | 0.3% | |
| perceptron | 1 | 0.3% | |
| ~moonshotai | 1 | 0.3% | |
| Kuaishou | 1 | 0.3% | |
| Upstage | 1 | 0.3% | |
| Writer | 1 | 0.3% | |
| deepcogito | 1 | 0.3% | |
| AI21 Labs | 1 | 0.3% | |
| Baidu | 1 | 0.3% | |
| Windsurf | 1 | 0.3% |
Providers are compared across multiple dimensions: average composite score across all models, best individual model score, stability rate (percentage of models in a stable state), 24-hour and 7-day momentum, model count, and availability of free models. Data is derived from hourly-updated rankings of 290+ tracked models.
The stability percentage represents the fraction of a provider's models that are in a "stable" ranking state. A higher percentage means the provider's models consistently maintain their positions over time, suggesting reliable and predictable performance across their lineup.
Market concentration is shown as each provider's share of the total tracked models. The visual bar chart and percentage table reveal how models are distributed among providers, helping identify whether the market is dominated by a few large players or spread across many competitors.