One fullstack platform for Compute, Inference, Fine-tuning, and RAG on Open Source Models.
| Model | Category | Type | Context | Strength |
|---|---|---|---|---|
 Llama 3.3 70B Instruct | General | Instruction | 128K | General assistant |
 Nemotron Nano 30B | Enterprise | Efficient | 128K | Enterprise AI |
 DeepSeek R1 | Reasoning | Advanced | 128K | Reasoning |
| DeepSeek R1 Distill 70B | Reasoning | Efficient | 128K | Distilled reasoning |
| DeepSeek V3 | General | Multilingual | 128K | General chat |
| DeepSeek V3.2 | General | Improved | 128K | General chat |
| DeepSeek V4 Flash | Fast | Low Latency | 128K | Fast inference |
| DeepSeek V4 Pro | Reasoning | High Performance | 256K | Advanced reasoning |
 GLM 4.7 | Chat | Multilingual | 128K | Conversational AI |
| GLM 5 | Reasoning | Multilingual | 128K | Reasoning |
 Kimi K2 Thinking | Reasoning | Long Context | 200K | Reasoning |
 Fara 7B | Enterprise | Lightweight | 32K | Enterprise AI |
 Minimax M2.5 | Multimodal | Chat | 128K | Vision + chat |
 Mistral 7B | Open | Efficient | 32K | Open-weight AI |
| Kimi K2 Instruct | Chat | Instruction | 200K | Conversational AI |
| Nemotron Nano Omni | Multimodal | Vision + Audio | 128K | Omni AI |
| Nemotron Super 120B | Enterprise | High Performance | 128K | Enterprise inference |
 GPT OSS 120B | Open | Large Model | 128K | Open reasoning |
 Qwen3 Max | General | Multilingual | 128K | General AI |
| Qwen3 Next 80B | Reasoning | Advanced | 256K | Reasoning |
| Qwen3 Coder Next | Coding | Advanced | 256K | Agentic coding |
| Qwen3 Coder Plus | Coding | Balanced | 128K | Software engineering |
| Qwen3 Coder 30B | Coding | Instruct | 128K | Code generation |
| Qwen3 Coder 480B | Coding | Large | 256K | Massive coding |
| Qwen3 Coder Flash | Coding | Fast | 128K | Fast coding |
| Qwen3 VL 235B Instruct | Vision | Instruction | 128K | Vision understanding |
| Qwen3 VL 235B Thinking | Vision | Reasoning | 128K | Vision reasoning |
| Qwen3 VL 30B | Vision | Efficient | 128K | Vision tasks |
| Qwen3 VL 8B | Vision | Lightweight | 128K | Light vision tasks |
| Qwen3 VL Flash | Vision | Fast | 128K | Fast multimodal |
| Qwen3 VL Plus | Vision | Advanced | 128K | Advanced multimodal |
| Qwen3.5 122B | General | Large | 128K | Multilingual AI |
| Qwen3.5 27B | General | Efficient | 128K | Efficient inference |
| Qwen3.5 35B | Reasoning | Balanced | 128K | Reasoning |
| Qwen3.5 397B | General | Massive | 256K | Large-scale AI |
| Qwen3.5 Flash | Fast | Efficient | 128K | Fast inference |
| Qwen3.5 Plus | General | Balanced | 128K | General AI |
| Qwen3.6 27B | General | Efficient | 128K | General inference |
| Qwen3.6 35B | Reasoning | Balanced | 128K | Reasoning |
| Qwen3.6 Max | General | Preview | 256K | High performance |
| Qwen3.6 Plus | General | Advanced | 128K | Advanced AI |
| Kimi K2.5 | Chat | Long Context | 200K | Long-context AI |
| Kimi K2.6 | Chat | Advanced | 200K | Conversational AI |
| GPT-4.1 | Coding | Reliable | 1M | Software engineering |
| GPT-4o | Multimodal | Omni | 128K | Vision + chat |
| GPT-4o Mini | Fast | Efficient | 128K | Lightweight multimodal |
 Gemini 2.5 Pro | Multimodal | Advanced | 1M | Long-context multimodal |
| Gemini 2.5 Flash | Fast | Efficient | 1M | Fast multimodal |
| Gemini 3 Flash | Fast | Preview | 1M | Next-gen flash AI |
| Gemini 3.1 Pro | General | Preview | 1M | Advanced reasoning |
 Tencent Hunyuan OCR | OCR | Document AI | 128K | OCR + extraction |
 Claude Haiku 4.5 | Fast | Efficient | 200K | Ultra-fast chat |
| Claude Opus 4.5 | Reasoning | Premium | 200K | High-end reasoning |
| Claude Opus 4.6 | Reasoning | Advanced | 200K | Advanced reasoning |
| Claude Opus 4.7 | Reasoning | Top Tier | 200K | Top-tier reasoning |
| Claude Sonnet 4.5 | Chat | Balanced | 200K | General assistant |
| Claude Sonnet 4.6 | Chat | Reasoning | 200K | Conversational reasoning |
| Qwen3 Plus | General | Balanced | 128K | General AI |
| GPT-5.4 | Reasoning | Frontier | 1M | Flagship reasoning |
| GPT-5.4 Mini | Reasoning | Efficient | 1M | Fast reasoning |
| GPT-5.4 Nano | Fast | Ultra Efficient | 1M | Ultra-low latency |
Run powerful AI models via simple APIs - no infrastructure required.
We handle routing, scaling, tuning, and reliability so your team can focus on building.
Need higher performance or predictable workloads?
Launch dedicated GPU instances with better latency, control, and consistent performance.
As demand grows, scale to high-performance infrastructure.
Move to bare metal and AI appliances for maximum performance and lower cost at scale.
From zero setup → dedicated compute → hyperscale infrastructure - all in one platform.