Every provider in one interface

Anthropic Claude models including the new Opus 4.5, the most intelligent model for complex tasks.

OpenAI delivers state-of-the-art multimodal and reasoning models including GPT-5.2, GPT-4o, o1, o3-mini, and GPT-4o-mini for fast, high-quality responses.

Google Gemini pairs long-context reasoning, multimodal understanding, and on-device options spanning Gemini Flash, Pro, and Nano tiers.

xAI Grok 4 delivers high-context reasoning, fast analysis, and multimodal understanding with optional reasoning traces and coding variants.

Groq delivers high-throughput, low-latency inference for open-weight models like LLaMA 3.3 and Mixtral — perfect for prototyping and production agents.

DeepSeek V3.2 and DeepSeek Coder provide deliberate reasoning, chain-of-thought, and competitive pricing for engineering and research teams with DeepSeek Sparse Attention.

Mistral provides high-quality open and hosted models — including Mistral Large 3, Devstral 2, and Mixtral — balancing speed, accuracy, and multilingual support.

Perplexity Sonar blends DeepSeek reasoning with live web retrieval, returning sourced answers for analysts and customer teams.

OpenRouter aggregates top community models including DeepSeek variants, LLaMA finetunes, and experimental releases, available through a single API.

Together AI offers performant hosted open models — including LLaMA 3.1 405B and Mixtral — with flexible pricing and finetuning options.

Fireworks AI accelerates open-source models like DeepSeek and LLaMA with tuned inference, convenient endpoints, and aggressive pricing.

Cohere Command models focus on multilingual accuracy, enterprise guardrails, and long-context tool use suited for knowledge work.

MiniMax specializes in compact, high-efficiency models optimized for agent workflows, coding, and tool use with exceptional cost-performance ratio.