The catalog grows to 190 models from 23 providers
As of February 2026, the idapt catalog spans 190 models from 23 providers. This release added over 170 of them, the biggest expansion we have shipped. The catalog has kept growing since; the models directory always shows the live count, every spec, and every price.
Sixteen providers join the catalog
| Provider | Flagships added | Known for |
|---|---|---|
| Mistral | Devstral, Codestral, Mistral Large 2512 | Coding, European multilingual |
| Meta | Llama 4 Maverick, Llama 4 Scout | Open weights, multimodal |
| NVIDIA | Nemotron Ultra 253B, Nemotron Super 49B | Fine-tuned Llama, high reasoning |
| Amazon | Nova Premier, Nova 2 Lite, Nova Micro | Enterprise, multimodal, cost-efficient |
| MiniMax | MiniMax M2.5 | Trillion-parameter MoE, long context |
| MoonshotAI | Kimi K2.5, Kimi K2 Thinking | 1T-parameter MoE, strong reasoning |
| Zhipu AI | GLM-5, GLM-4.7 | Chinese-English bilingual |
| ByteDance Seed | Seed-1.6, Seed-1.6 Flash | Fast, capable bilingual |
| StepFun | Step-3.5 Flash | Efficient reasoning |
| Baidu | ERNIE 4.5 300B, ERNIE 4.5 VL | Chinese language, multimodal |
| Cohere | Command A | Enterprise RAG |
| NousResearch | Hermes 4 405B, Hermes 4 70B | Fine-tuned open weights, agentic |
| Arcee AI | Virtuoso Large, Maestro Reasoning | Model merging, compact reasoning |
| AionLabs | Aion-1.0, Aion-1.0 Mini | Extended reasoning |
| Inception AI | Mercury, Mercury Coder | Diffusion LLMs, fast |
| TNG Tech | DeepSeek R1T2 Chimera | Model merging |
The existing providers, expanded
OpenAI: the full GPT-5 family. GPT-5.2, GPT-5, GPT-5 Mini, and GPT-5 Nano are live, with the GPT-4.1 series, the reasoning models o3 and o4-mini, and the OSS and legacy models back to GPT-3.5.
Anthropic: every current Claude. Claude Opus 4.6, Claude Sonnet 4.6, and Claude Sonnet 4.5, alongside the full Claude 4 generation and Claude 3.7 Sonnet with extended thinking.
xAI: Grok 4 and 4.1. Grok 4 and Grok 4.1 Fast join Grok Code Fast and the Grok 3 family.
Google: Gemini 3 plus the full Gemma line. Gemini 3.1 Pro Preview, Gemini 3 Flash, Gemini 2.5 Pro and Flash, and the complete Gemma 3 series (4B, 12B, 27B).
DeepSeek: R1 and V3.2. DeepSeek R1 and DeepSeek V3.2, with R1-0528, the R1 distills (Qwen 32B, Llama 70B), and V3.1.
Qwen: 20+ models including Qwen 3 Coder. Qwen 3 Coder Plus, Qwen 3 235B, QwQ-32B, and the vision models.
Every model, one workspace
Every catalog model works in the same chat: switch mid-conversation, run one prompt across several models in parallel tabs, and keep your files and history in place. Each model has a detail page with benchmarks, pricing, and capabilities, and any two can be compared side by side.
Browse the live catalog in the models directory; it has grown well past this announcement's count.
Found this helpful? Share it: