# The AI labs and platforms we work with > We build with models and tools from every major lab, and we teach your teams to use them in their daily work. We are independent: we choose per task and keep your option to switch. - Page: https://www.ninalabs.ai/platforms/ - Publisher: Nina Labs AI LLC ## Anthropic and OpenAI Most of our enablement work runs on Claude and ChatGPT. We set them up for the whole company, connect them to your tools and train each team on its own tasks. ### Anthropic Products: Claude, Claude Code, Claude Agent SDK, Claude Managed Agents. We roll out Claude to whole companies and build production agents on the Claude Agent SDK. Engineering teams get Claude Code set up on their own repositories. - Claude Enterprise setup: SSO, roles, data controls and connectors to your tools - Projects and Agent Skills for each team's repeat tasks - Claude Code for engineers: CLAUDE.md files, hooks, permissions, review rules and CI - Production agents on the Claude Agent SDK or Claude Managed Agents - Claude through the Claude API, Amazon Bedrock, Google Cloud or Microsoft Foundry ### OpenAI Products: ChatGPT, Codex, Agents SDK, Responses API. We roll out ChatGPT Enterprise and Codex, train each team on its real work, and build agents on the OpenAI Agents SDK and the Responses API. - ChatGPT Enterprise setup: SSO, workspace roles, data controls and connectors - Shared projects and workflows for each team's repeat tasks - Codex for engineers: cloud tasks, CLI, IDE extension, AGENTS.md files and code review - Production agents on the OpenAI Agents SDK and the Responses API, including realtime voice - OpenAI models through the OpenAI API, Microsoft Foundry or Amazon Bedrock ## Agent platforms on your cloud - **Google Cloud** (Vertex AI · Gemini Enterprise · ADK): Agents and workflows on Vertex AI, now the Gemini Enterprise Agent Platform, with the Agent Development Kit, Gemini models and Gemini Enterprise for teams. - **Microsoft** (Microsoft Foundry · Copilot Studio · Agent Framework): Agents on Microsoft Foundry and the Microsoft Agent Framework, Copilot Studio for business teams, and Entra Agent ID for agent identity. - **Amazon Web Services** (Amazon Bedrock · AgentCore · Strands): Claude, OpenAI and open-weight models on Amazon Bedrock, with Bedrock AgentCore for runtime, memory, identity and observability. ## More labs - **Google DeepMind** (Gemini · Gemma): Gemini models for long-context and multimodal work, and Gemma for open-weight deployments. - **Meta** (Llama): Llama open-weight models for private deployments and fine-tuning on your own data. - **Mistral AI** (Mistral Large · Codestral): European models with on-premises options, for teams with strict data residency rules. - **Perplexity** (Sonar API · Enterprise): Search-grounded answers with citations, for research agents and current-events questions. - **DeepSeek** (DeepSeek models): Open-weight reasoning and coding models with permissive licences, run on infrastructure you control. - **Z.ai** (GLM models): GLM open-weight models from Z.ai for coding and agent tasks, hosted where your data rules allow. - **Alibaba Qwen** (Qwen models): A wide range of Qwen open-weight models, from small on-device models to large coding models. - **Moonshot AI** (Kimi models): Kimi models for long-context and agentic coding work, deployed on infrastructure you choose. - **MiniMax** (MiniMax models): Open-weight text and speech models, used where cost per token or latency decides the choice. - **Cohere** (Command · Embed · Rerank): Retrieval models for search over company knowledge, with private deployment options. - **NVIDIA** (NIM · Nemotron): NIM microservices and Nemotron models for high-throughput serving on your own GPUs. - **Grok** (Grok models): Grok models through the API, evaluated against the other labs on your own cases. ## How we choose a model - **Measured on your cases:** We run the same eval set against several models and compare quality, not marketing claims. - **Cost and latency:** We price each option per completed task, including retries, and check the response time your users need. - **Data and residency:** We match each model to your data rules: vendor API, your cloud, or open weights on your own hardware. - **Room to switch:** We build on open protocols such as MCP and keep prompts and evals portable, so a model change is a configuration change. ## Full stack - **Frontier labs:** Anthropic Claude, OpenAI GPT, Google Gemini, Meta, Mistral AI, Perplexity, Cohere, Grok - **Open-weight models:** DeepSeek, Z.ai GLM, Qwen, Kimi, MiniMax, Llama, Gemma, NVIDIA Nemotron - **Agent platforms:** Claude Agent SDK, OpenAI Agents SDK, Google ADK, Vertex AI, Microsoft Foundry, Amazon Bedrock, LangGraph, LlamaIndex - **Coding agents:** Claude Code, OpenAI Codex, GitHub Copilot, Cursor, Kiro, Google Antigravity, Windsurf - **Protocols:** Model Context Protocol, Agent2Agent (A2A), AGENTS.md, Agent Skills - **Clouds and runtimes:** AWS, Microsoft Azure, Google Cloud, NVIDIA NIM, Hugging Face, vLLM, Ollama, Your own hardware - **Orchestration:** Temporal, n8n, AWS Step Functions, Azure Durable Functions - **Observability and evals:** OpenTelemetry, Langfuse, LangSmith, Braintrust, Arize Phoenix, Promptfoo ## Frequently asked questions ### Are you an official partner of these companies? We are independent. We use these labs' products and APIs for our clients and are not affiliated with them, apart from the Microsoft for Startups program we belong to. The logos on this site identify the products we work with. ### Can you train our teams on Claude and ChatGPT? Yes. We set up Claude Enterprise, ChatGPT Enterprise or both, write the data rules with your security team, and run hands-on workshops for each team on its own tasks. Engineers get Claude Code and Codex set up on their repositories. ### Do you work with Google Vertex AI? Yes. Vertex AI is now part of the Gemini Enterprise Agent Platform. We build agents there with the Agent Development Kit and Gemini models, and we also run Claude and open-weight models on Google Cloud. ### Can we use DeepSeek, Qwen, GLM or Kimi models safely? Their open-weight models can run on infrastructure you control, such as your own cloud account or GPUs, so no data goes to the model maker. We check each model's licence and run your evals before we recommend it. ### What if a better model comes out next month? That happens often. We keep prompts, tools and eval sets portable and build on open protocols such as MCP. A model change is then a configuration change plus an eval run, not a rebuild. ## Contact Talk to an engineer through the [contact form](https://www.ninalabs.ai/contact/) or by email at contact@ninalabs.ai.