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 | Xiaomi | 3 | 84.2 | MiMo-V2-Omni(85.0) | #22 | +10.7 | +182.7 | 0% | 0 |
| 2 | ByteDance | 5 | 80.5 | Seed-2.0-Lite(85.0) | #25 | +4.2 | +7.8 | 40% | 0 |
| 3 | xAI | 10 | 78.8 | Grok 4.1 Fast(86.9) | #16 | -3.8 | +5.2 | 30% | 0 |
| 4 | Kuaishou | 1 | 77.4 | KAT-Coder-Pro V1(77.4) | #90 | -4.0 | +16.0 | 0% | 0 |
| 5 | Anthropic | 13 | 77.3 | Claude Opus 4.6(92.1) | #6 | -3.6 | -2.3 | 54% | 0 |
| 6 | Cursor | 2 | 76.4 | Composer 2(76.4) | #101 | +5.0 | -0.5 | 100% | 0 |
| 7 | StepFun | 2 | 75.7 | Step 3.5 Flash (free)(78.2) | #86 | -12.0 | -3.0 | 50% | 1 |
| 8 | Meituan | 1 | 72.8 | LongCat Flash Chat(72.8) | #130 | -7.0 | +16.0 | 0% | 0 |
| 9 | OpenAI | 60 | 72.5 | GPT-5.4 Pro(94.0) | #1 | -0.2 | +10.4 | 42% | 2 |
| 10 | Upstage | 1 | 72.5 | Solar Pro 3(72.5) | #139 | +14.0 | -2.0 | 100% | 0 |
| 11 | Tencent | 1 | 72.3 | Hunyuan A13B Instruct(72.3) | #141 | -16.0 | -27.0 | 0% | 0 |
| 12 | MiniMax | 8 | 72.3 | MiniMax M2.5 (free)(83.4) | #51 | +3.1 | +38.4 | 0% | 1 |
| 13 | 23 | 71.9 | Gemini 3 Pro Preview(90.3) | #10 | +2.7 | -0.6 | 57% | 5 | |
| 14 | Moonshot AI | 4 | 71.5 | Kimi K2.5(85.0) | #31 | -7.0 | -1.5 | 50% | 0 |
| 15 | DeepSeek | 11 | 71.5 | R1 0528(77.7) | #89 | +4.5 | -3.5 | 18% | 0 |
| 16 | AI21 Labs | 1 | 71.2 | Jamba Large 1.7(71.2) | #150 | -14.0 | -7.0 | 0% | 0 |
| 17 | Inception | 3 | 70.6 | Mercury 2(81.3) | #70 | -2.0 | +4.0 | 67% | 0 |
| 18 | NVIDIA | 11 | 70.6 | Nemotron 3 Super (free)(84.1) | #48 | +3.3 | -1.4 | 36% | 4 |
| 19 | Alibaba | 51 | 70.2 | Qwen3.5 Plus 2026-02-15(85.0) | #30 | -0.7 | -0.4 | 45% | 3 |
| 20 | Baidu | 5 | 68.6 | ERNIE 4.5 VL 28B A3B(75.0) | #110 | +11.2 | +3.6 | 40% | 0 |
| 21 | deepcogito | 1 | 66.7 | Cogito v2.1 671B(66.7) | #174 | +15.0 | +6.0 | 0% | 0 |
| 22 | essentialai | 1 | 64.8 | Rnj 1 Instruct(64.8) | #188 | +14.0 | -5.0 | 100% | 0 |
| 23 | arcee-ai | 7 | 64.7 | Trinity Mini(82.4) | #59 | -1.7 | +0.6 | 43% | 2 |
| 24 | Writer | 1 | 64.7 | Palmyra X5(64.7) | #190 | +6.0 | 0 | 100% | 0 |
| 25 | Perplexity | 5 | 63.8 | Sonar Pro Search(85.0) | #39 | -0.6 | +4.6 | 40% | 0 |
| 26 | Amazon | 5 | 63.6 | Nova Premier 1.0(77.8) | #88 | 0 | -5.8 | 60% | 0 |
| 27 | aion-labs | 3 | 60.8 | Aion-2.0(69.2) | #159 | -8.0 | +4.0 | 0% | 0 |
| 28 | Mistral AI | 25 | 60.6 | Mistral Small 4(79.4) | #79 | -1.5 | +7.4 | 44% | 1 |
| 29 | Allen AI | 4 | 60.1 | Olmo 3 32B Think(66.3) | #175 | +2.5 | -3.5 | 50% | 0 |
| 30 | IBM | 1 | 55.1 | Granite 4.0 Micro(55.1) | #245 | +2.0 | +2.0 | 100% | 0 |
| 31 | Liquid AI | 5 | 54.4 | LFM2.5-1.2B-Thinking (free)(59.0) | #226 | 0 | -3.8 | 40% | 2 |
| 32 | Cohere | 4 | 50.1 | Command A(60.0) | #220 | -4.0 | 0 | 100% | 0 |
| 33 | Meta | 14 | 49.9 | Llama 4 Maverick(76.7) | #99 | -1.9 | -4.1 | 64% | 2 |
| 34 | Windsurf | 1 | 49.2 | SWE-1.5(49.2) | #263 | -2.0 | +5.0 | 100% | 0 |
| 35 | eleutherai | 1 | 47.5 | Llemma 7b(47.5) | #266 | 0 | -1.0 | 100% | 0 |
| 36 | Microsoft | 2 | 45.9 | Phi 4(59.6) | #223 | +0.5 | +4.0 | 50% | 0 |
| 37 | Vercel | 1 | 38.8 | autofixer-01(38.8) | #287 | -1.0 | +2.0 | 100% | 0 |
| 38 | Inflection | 2 | 36.8 | Inflection 3 Pi(36.8) | #293 | +1.0 | -3.0 | 100% | 0 |
| 39 | JetBrains | 1 | 32.6 | Mellum(32.6) | #297 | +2.0 | -1.0 | 100% | 0 |
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 | 60 | 20.0% | |
| Alibaba | 51 | 17.0% | |
| Mistral AI | 25 | 8.3% | |
| 23 | 7.7% | ||
| Meta | 14 | 4.7% | |
| Anthropic | 13 | 4.3% | |
| DeepSeek | 11 | 3.7% | |
| NVIDIA | 11 | 3.7% | |
| xAI | 10 | 3.3% | |
| MiniMax | 8 | 2.7% | |
| arcee-ai | 7 | 2.3% | |
| ByteDance | 5 | 1.7% | |
| Baidu | 5 | 1.7% | |
| Perplexity | 5 | 1.7% | |
| Amazon | 5 | 1.7% | |
| Liquid AI | 5 | 1.7% | |
| Moonshot AI | 4 | 1.3% | |
| Allen AI | 4 | 1.3% | |
| Cohere | 4 | 1.3% | |
| Xiaomi | 3 | 1.0% | |
| Inception | 3 | 1.0% | |
| aion-labs | 3 | 1.0% | |
| Cursor | 2 | 0.7% | |
| StepFun | 2 | 0.7% | |
| Microsoft | 2 | 0.7% | |
| Inflection | 2 | 0.7% | |
| Kuaishou | 1 | 0.3% | |
| Meituan | 1 | 0.3% | |
| Upstage | 1 | 0.3% | |
| Tencent | 1 | 0.3% | |
| AI21 Labs | 1 | 0.3% | |
| deepcogito | 1 | 0.3% | |
| essentialai | 1 | 0.3% | |
| Writer | 1 | 0.3% | |
| IBM | 1 | 0.3% | |
| Windsurf | 1 | 0.3% | |
| eleutherai | 1 | 0.3% | |
| Vercel | 1 | 0.3% | |
| JetBrains | 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.