PROVIDER PORTFOLIOS
Understand the portfolio.
Case aliases are grouped; distinct brands stay separate. Sample size and missing prices remain visible.
Recent snapshotData snapshot cf-1.3
| Provider | Models | Median index | Median listed cost | Price unknown |
|---|---|---|---|---|
| Unknown | 230 | 12.3 | $1 | 167 / 230 |
| OpenAI | 30 | 53.6 | $12 | 3 / 30 |
| Anthropic | 24 | 60.3 | $30 | 7 / 24 |
| 19 | 49.5 | $4.5 | 6 / 19 | |
| xAI | 11 | 41.7 | $3.75 | 3 / 11 |
| Meta | 10 | 55.4 | $5.5 | 5 / 10 |
| DeepSeek | 9 | 46.1 | $1.5 | 2 / 9 |
| Alibaba Cloud / Qwen Team | 7 | 64.0 | $3.825 | 5 / 7 |
| Alibaba | 7 | 49.0 | $1.65 | 3 / 7 |
| Qwen | 7 | 40.3 | $5 | 4 / 7 |
| InternScience | 6 | 53.4 | — | 6 / 6 |
| NVIDIA | 6 | 39.0 | $1.2 | 4 / 6 |
| zai-org | 5 | 52.4 | $5 | 1 / 5 |
| Xiaomi | 5 | 47.5 | $0.863 | 3 / 5 |
| Moonshot AI | 4 | 40.9 | $3.6 | 0 / 4 |
| Zhipu AI | 4 | 53.0 | $3.25 | 2 / 4 |
| InclusionAI | 4 | 50.8 | $0.24 | 2 / 4 |
| moonshotai | 3 | 59.3 | $4.95 | 0 / 3 |
| Reflection AI | 3 | 60.2 | $1.6 | 0 / 3 |
| Ornith AI | 3 | 29.7 | — | 3 / 3 |
| minimax | 3 | 45.2 | $1.5 | 0 / 3 |
| Tencent | 3 | 43.1 | — | 3 / 3 |
| Mistral | 3 | 23.8 | $2.4 | 0 / 3 |
| xiaomi | 2 | 52.2 | $0.905 | 0 / 2 |
| Thinking Machines Lab | 2 | 63.0 | $4.285 | 0 / 2 |
| StepFun | 2 | 57.6 | $2.525 | 0 / 2 |
| Microsoft | 2 | 25.8 | — | 2 / 2 |
| Dots Studio | 1 | 70.5 | — | 1 / 1 |
| Z.AI | 1 | 65.0 | $5.2 | 0 / 1 |
| Recursive | 1 | 64.2 | — | 1 / 1 |
| MiniMax | 1 | 55.8 | $0.75 | 0 / 1 |
| Prism ML | 1 | 54.6 | — | 1 / 1 |
| Arcee AI | 1 | 54.0 | $1.15 | 0 / 1 |
| Upstage | 1 | 53.0 | $0.75 | 0 / 1 |
| Apodex | 1 | 52.7 | — | 1 / 1 |
| Mistral AI | 1 | 45.7 | $2.8 | 0 / 1 |
| Multiverse Computing | 1 | 39.1 | $2.4 | 0 / 1 |
| SparkLLM | 1 | 35.9 | — | 1 / 1 |
| LiquidAI | 1 | 34.6 | $1.1 | 0 / 1 |
| IBM | 1 | 31.5 | — | 1 / 1 |
| Cohere | 1 | 30.4 | $12.5 | 0 / 1 |
| Liquid AI | 1 | 29.4 | — | 1 / 1 |
| Amazon | 1 | 27.0 | — | 1 / 1 |
| Poolside | 1 | 26.3 | — | 1 / 1 |
| Inception | 1 | 24.7 | $0.19 | 0 / 1 |
Listed dataset prices; provider terms and dates need verification. Zero without explicit free-price evidence is treated as unknown. Cost combines input and output prices per million tokens. Different product samples prevent a simple provider quality ranking.