providers
Every provider. One account.
9 providers, 410 active models, one interface. Every provider listed here has live models you can use today.
OpenAI develops multimodal and reasoning models including GPT-6 Astra, the GPT-5.6 family, GPT-5.5, and the o-series.
GPT-6 Astra · GPT-5.6 Luna · GPT-5.6 Luna Pro
Mistral provides high-quality open and hosted models, including Mistral Large 3, Devstral 2 and Mixtral, balancing speed, accuracy and multilingual support.
Devstral 2 2512 · Mistral Large 3 2512 · Codestral 2508
Google Gemini pairs long-context reasoning, multimodal understanding, image-native generation, and open Gemma families spanning Gemini 3.8, Gemini 2.5, and Gemma 4 tiers.
Gemini 3.8 Flash · Gemini 3.5 Flash Lite · Gemini 3.6 Flash
Anthropic Claude models including Claude Fable 5.1 and Claude Opus 5 for long-running, complex tasks.
Claude Fable 5.1 · Claude Opus 5 · Claude Sonnet 5
Cohere Command models focus on multilingual accuracy, enterprise guardrails, long-context tool use, and the newer reasoning, vision, and translation specialisations.
Command A Reasoning (Aug 2025) · Command A Vision (Jul 2025) · Command A (Mar 2025)
xAI Grok delivers high-context reasoning, fast analysis, multimodal understanding, and tool-heavy agent workflows across Grok 4.3 and the active Grok 4.20 variants.
Grok 4.5 · Grok 4.3 · Grok 4.20 (Fast)
DeepSeek V4 and DeepSeek Reasoner provide deliberate reasoning, coding, and competitive pricing for engineering and research teams with long-context support.
DeepSeek V4 Flash · DeepSeek V4 Pro · DeepSeek Reasoner
Perplexity Sonar blends DeepSeek reasoning with live web retrieval, returning sourced answers for analysts and customer teams.
Sonar Reasoning Pro · Sonar Deep Research · Sonar Pro
Together AI offers performant hosted open models, including LLaMA 3.1 405B and Mixtral, with flexible pricing and finetuning options.
DeepSeek R1 0528 · DeepSeek V3 · Qwen 2.5 7B Instruct Turbo