Last updated: 2026-10-27
NX is Nexplora's own model and the default. The model picker lists 59 models (40 for chat and code, 19 for images and video) from NX and 12 providers. On the desktop app you can also use your ChatGPT or Claude plan: you sign in with them directly, and Nexplora never sees that token. Or add your own API keys in Settings. Which models you can run depends on your plan and the keys you connect.
| Model | Provider | Best for | Context |
|---|---|---|---|
| NX (default) | Nexplora | Nexplora's managed default | 256K tokens |
| Fable 5 | Anthropic | Long-running agents and deepest reasoning | 1M tokens |
| Opus 5 | Anthropic | Complex coding and enterprise work | 1M tokens |
| Sonnet 5 | Anthropic | Fast, balanced intelligence | 1M tokens |
| Haiku 4.5 | Anthropic | Lowest-latency Claude | 200K tokens |
| GPT-5.6 Sol | OpenAI | Frontier capability | 1.05M tokens |
| GPT-5.6 Terra | OpenAI | Balanced intelligence and cost | 1.05M tokens |
| GPT-5.6 Luna | OpenAI | Fast, high-volume work | 1.05M tokens |
| Gemini 3.6 Flash | Fast frontier agentic loops | 1.048576M tokens | |
| Gemini 3.5 Flash | Sustained coding workflows | 1.048576M tokens | |
| Gemini 3.5 Flash-Lite | Lowest-cost high throughput | 1.048576M tokens | |
| Gemini 3.1 Pro Preview | Deep software reasoning | 1.048576M tokens | |
| Grok 4.5 | xAI | Agentic coding and engineering | 500K tokens |
| DeepSeek V3 | DeepSeek | DeepSeek's flagship chat model | 128K tokens |
| DeepSeek R1 | DeepSeek | DeepSeek's deep-reasoning model | 128K tokens |
| Kimi K2 | Moonshot AI | Moonshot's agentic long-context model | 256K tokens |
| Qwen Max | Alibaba Qwen | Alibaba's flagship Qwen model | 128K tokens |
| Qwen3 Coder | Alibaba Qwen | Alibaba's coding-tuned Qwen | 256K tokens |
| Llama 4 Scout | Groq | Meta Llama 4 Scout on Groq — fastest inference | 131K tokens |
| Llama 4 Maverick | Groq | Meta Llama 4 Maverick on Groq — deep reasoning, fast | 131K tokens |
| DeepSeek R1 Distill | Groq | DeepSeek R1 70B distill on Groq — reasoning, instant | 128K tokens |
| Mistral Large 3 | Mistral AI | Mistral's flagship reasoning + code model | 256K tokens |
| Mistral Small | Mistral AI | Mistral's efficient business model | 128K tokens |
| Codestral | Mistral AI | Mistral's coding specialist | 256K tokens |
| Command A | Cohere | Cohere's enterprise RAG + tool-calling model | 256K tokens |
| Command R+ | Cohere | Cohere's large model — long context, deep reasoning | 128K tokens |
| Sonar Pro | Perplexity | Perplexity's search-grounded deep reasoning | 200K tokens |
| Sonar | Perplexity | Perplexity's fast search-grounded model | 128K tokens |
| MiniMax M3 | OpenRouter | MiniMax's long-context reasoning model | 1M tokens |
| Nemotron 3 Ultra | OpenRouter | NVIDIA's largest reasoning model | 128K tokens |
| Nemotron 3 Omni | OpenRouter | NVIDIA's multimodal reasoning model | 128K tokens |
| Llama 4 Scout | OpenRouter | Meta's multimodal long-context model | 1M tokens |
| Mistral Small | OpenRouter | Mistral's efficient business model | 128K tokens |
| Mistral Large 3 | OpenRouter | Mistral's flagship reasoning + code model | 256K tokens |
| Command A | OpenRouter | Cohere's enterprise RAG + tool model | 256K tokens |
| Sonar Pro | OpenRouter | Perplexity's search-grounded reasoning model | 200K tokens |
| Nova Pro | OpenRouter | Amazon's multimodal business model | 300K tokens |
| DBRX Next | OpenRouter | Databricks' agentic engineering model | 128K tokens |
| Ling 3.0 Flash | OpenRouter | InclusionAI's fast open model | 128K tokens |
| Poolside Laguna | OpenRouter | Poolside's software-engineering model | 128K tokens |
| Model | Makes | Notes |
|---|---|---|
| Gemini 3 Pro Image | Images | Google — Nexplora's default image model, top quality |
| GPT Image 1 Mini | Images | OpenAI — cheaper, fast |
| Nano-Banana | Images | Google — fast, high fidelity (your key) |
| Qwen-Image | Images | Alibaba — best value, superb text (your fal key) |
| Qwen-Image Max | Images | Alibaba — most realistic Qwen tier (your fal key) |
| FLUX 1.1 Pro | Images | Black Forest Labs — best quality (your key) |
| FLUX dev | Images | Black Forest Labs — fast, open (your key) |
| Ideogram 3.0 | Images | Ideogram — best in-image text (your key) |
| Ideogram 2.0 | Images | Ideogram — balanced (your key) |
| Seedream 5.0 Pro | Images | ByteDance — flagship, 2K/3K + layouts (your fal key) |
| Seedream 4 | Images | ByteDance — fast, reliable (your fal key) |
| Seedream 3 | Images | ByteDance — stable (your fal key) |
| Stable Image Core | Images | Stability AI — fast, reliable (your key) |
| Stable Image Ultra | Images | Stability AI — top quality (your key) |
| Veo 3.1 Fast | Video | Google — 8s video with sound, on your Nexplora credits |
| Veo 3.1 (high quality) | Video | Google — highest-quality 8s video with sound, on your Nexplora credits |
| Wan 2.7 | Video | Alibaba — text→video, 720p/1080p (your fal key) |
| HeyGen Avatar | Video | HeyGen — AI presenter speaks your script (your key) |
| Runway Gen-4.5 | Video | Runway — image→video, needs a reference image (your key) |
Add your own API keys in Settings, or on the desktop app sign in with your ChatGPT or Claude plan; Nexplora never sees that sign-in token. Keys you add are encrypted at rest and never shown back. Connecting a provider does not bypass Nexplora's permissions or approvals.
Nexplora (nexplora.ai) is an AI workspace for running a business. Specialist agents for sales, marketing, finance, support and operations do the work in the tools you connect, research shows the sources it used, and CodeChat writes and fixes code. Anything that sends, posts, pays or deletes waits for your approval.
Yes. Nexplora (nexplora.ai) is an AI workspace for running a business, where specialist agents for sales, marketing, finance, support and operations work in your connected tools with your approval, on 50+ models: 40 for chat and code from NX and 12 providers, including OpenAI, Anthropic and Google, plus 19 for images and video. NX is the default and you can switch at any time. Which models you can run depends on your plan and the keys you connect.
Yes. In Nexplora you pick the model per conversation from one picker that lists OpenAI GPT, Anthropic Claude and Google Gemini models next to NX. Your company memory, connected tools and approvals stay the same whichever model answers.
Yes. Nexplora's Council lets up to 4 more connected models work beside the lead model on the same task, in Chat and CodeChat. At the start of each turn the lead assigns roles, such as who plans, who builds and who reviews.
Yes. Nexplora lets you add your own API keys in Settings, and on the desktop app you can sign in with your ChatGPT or Claude plan; Nexplora never sees that sign-in token. Keys are encrypted at rest and never shown back. Bringing your own key changes who bills for model usage; it does not bypass Nexplora's plan limits, permissions or approvals.
Multi-model AI apps compared · Nexplora vs TypingMind · Nexplora vs Poe
Privacy PolicyTerms of ServiceData deletionSecurityContactLinkedInXGitHub