| Model | Type | Price, $/1M | Context | Released |
|---|---|---|---|---|
| MAI Code 1 Flash | text | $1.69 | 256K | Jun 2026 |
| Phi 4 Multimodal дообучена выполнять инструкции | text | $0.08 | 128K | Jul 2025 |
| WizardLM-2 8x22B | text | $0.48 | 66K | Apr 2025 |
| Phi 4 Mini 3.8B deprecated | text | $0.13 | 131K | Feb 2025 |
| MAI DS R1 | text | $2.36 | 66K | Jan 2025 |
| Phi 4 | text | $0.08 | 128K | Dec 2024 |
| Phi 4 Mini | text | $0.13 | 128K | Dec 2024 |
| Phi 4 Mini Reasoning | text | $0.13 | 128K | Dec 2024 |
| Phi 4 Multimodal phi-4-multimodal | text | $0.14 | 128K | Dec 2024 |
| Phi 4 Reasoning | text | $0.22 | 32K | Dec 2024 |
| Phi 4 Reasoning Plus | text | $0.22 | 32K | Dec 2024 |
| Phi-3.5-mini instruct (128k) | text | $0.23 | 128K | Aug 2024 |
| Phi-3.5-MoE instruct (128k) | text | $0.28 | 128K | Aug 2024 |
| Phi-3.5-vision instruct (128k) | text | $0.23 | 128K | Aug 2024 |
| E5 Multi-Lingual Large Embeddings 0.6B | text | $0.11 | 1K | Jun 2024 |
| Phi-3-medium instruct (128k) | text | $0.3 | 128K | Apr 2024 |
| Phi-3-medium instruct (4k) | text | $0.3 | 4K | Apr 2024 |
| Phi-3-mini instruct (128k) | text | $0.1 | 128K | Apr 2024 |
| Phi-3-mini instruct (4k) | text | $0.23 | 4K | Apr 2024 |
| Phi-3-small instruct (128k) | text | $0.26 | 128K | Apr 2024 |
| Phi-3-small instruct (8k) | text | $0.26 | 8K | Apr 2024 |
| Mai Code 1 Flash | text | $1.69 | 128K | — |
| Mai Code 1 Flash Internal | text | $1.69 | 128K | — |
| Mai Image 2.5 | image | — | — | — |
| Mai Image 2.5 Flash | image | — | — | — |
| Mai Image 2e | image | — | — | — |
| Orca Mini | text | — | 4K | — |
| Phi 2 3B | text | $0.1 | 2K | — |
| Phi 3 Vision 128K Instruct | text | $0.2 | 32K | — |
| Phind Code Llama 34B Python V1 | text | $0.9 | 16K | — |
| Phind Code Llama 34B V1 | text | $0.9 | 16K | — |
| Phind Code Llama 34B V2 | text | $0.9 | 16K | — |
| Prebuilt Document | ocr | — | — | — |
| Prebuilt Layout | ocr | — | — | — |
| Prebuilt Read | ocr | — | — | — |
A blended price, on a scheme of three input tokens per one output token. It is taken from the developer itself, if they sell this model; if not, from the cheapest offer among providers, as a pair from one price list. Base per-token tier only. A dash means the model has no per-token price in open sources, not that it is free.