AI Agent Tools With GPT/OpenAI Support
48 of the 94 tools in our index can run on GPT/OpenAI models — 4 exclusively, 15 offering it as one of a small vendor-picked set of frontier models, and 30 reaching it through a broader model-agnostic or bring-your-own-key layer alongside dozens of other providers.
On this page
"Supports GPT" means something different depending on which of these 48 tools you're looking at, and the difference matters if you're choosing one to integrate. We read every listing's own sourced `modelLLM` fact — and, where that field alone was silent, its `whatItDoes` or `integrations` text — rather than trusting the `gpt` tag alone. That check also had to rule out several tools that only LOOK like a match on a keyword search: Ada, Sierra and GC AI each name "ChatGPT" or "OpenAI" somewhere in their own listing, but as a caller integration or a founder's biography, never as the model actually powering their own agent, so none of the three is included here.
**GPT-only (4)** — no other model vendor is offered at all: [OpenAI Realtime API](/openai-realtime-api), [Synthflow AI](/synthflow-ai), [Microsoft Fabric Data Agent](/microsoft-fabric-data-agent) and [Unify](/unify), whose own listing names no model but GPT anywhere in its research.
**GPT among a short, vendor-picked list (15)** — the vendor offers GPT alongside a small, fixed set of other frontier models, usually Claude and/or Gemini, not an open provider layer: [Cursor](/cursor), [GitHub Copilot](/github-copilot), [Devin](/devin), [Windsurf](/windsurf), [Relevance AI](/relevance-ai), [Clay](/clay), [Perplexity](/perplexity), [Salesforce Agentforce](/salesforce-agentforce), [PolyAI](/polyai), [CodeRabbit](/coderabbit), [Glean](/glean), [Genspark](/genspark), [Zendesk AI Agents](/zendesk-ai-agents), [OpenAI Codex CLI](/codex-cli) and [Julius AI](/julius-ai).
**GPT via a broad model-agnostic or BYOK layer (30)** — GPT is one of many, often dozens, of named providers reachable through a generic bring-your-own-key or LiteLLM-style integration, not a vendor-curated shortlist: [n8n](/n8n), [Microsoft AutoGen](/microsoft-autogen), [Microsoft Agent Framework](/microsoft-agent-framework), [OpenAI Agents SDK](/openai-agents-sdk), [Agno](/agno), [Dify](/dify), [GPT Researcher](/gpt-researcher), [ElevenLabs Conversational AI](/elevenlabs-conversational-ai), [Juggler](/juggler), [OpenCode](/opencode), [Zot](/zot), [Warp](/warp), [Zed](/zed), [Vecbase](/vecbase), [Google ADK](/google-adk), [Botpress](/botpress), [Deepgram Voice Agent API](/deepgram-voice-agent-api), [Junie](/junie), [OpenClaw](/openclaw), [LangChain](/langchain), [Cline](/cline), [Aider](/aider), [Pydantic AI](/pydantic-ai), [OpenHands](/openhands), [LiveKit Agents](/livekit-agents), [Mastra](/mastra), [Cartesia](/cartesia), [CrewAI](/crewai), [Goose](/goose) and [LangGraph](/langgraph).
[Coding agents](/category/coding-agents) lead the count at 16 of 25 — GPT access is close to the category norm there — followed by [frameworks](/category/frameworks) at 10 of 14, [voice agents](/category/voice-agents) at 7 of 12, [agent platforms](/category/agent-platforms) at 9 of 16 and [research and data agents](/category/research-agents) at 5 of 11. [Support agents](/category/support-agents) has just 1 entry, Zendesk AI Agents — GPT access is still the exception, not the rule, in that category.
Why this collection
A tool earns a place here only if its own Published, sourced listing's `attributes.modelLLM` field — or, where that field alone was silent, its `whatItDoes`/`integrations`/FAQ text — names OpenAI or GPT specifically as a model the tool itself runs on. A generic "model-agnostic" or "bring your own model" claim with no provider named anywhere in the listing does not qualify on its own, and neither does a mention of "ChatGPT" or "OpenAI" that describes something other than the tool's own underlying model — a caller integration, a founder's background, or a named competitor. Every entry is additionally sorted into one of three tiers: GPT-only (no other vendor's models are offered at all), GPT among a small vendor-picked set of frontier models, or GPT reachable through a broad model-agnostic/BYOK layer alongside many other named providers. Verified directly against each listing's own already-researched, gate-passed fields, not inferred from the `gpt` tag alone. Re-verified against the live 89-listing corpus on 2026-08-18: removed Vapi (no named model provider anywhere in its listing history) and LlamaIndex (named OpenAI as a model provider until an unrelated 2026-08-16 edit dropped that line; flagged to listing for restoration rather than asserted here from memory), and added Goose (created 2026-08-17, names OpenAI explicitly). Net 50 -> 49.
48 agents in this collection
Best for Engineering teams building a voice agent from scratch who want to go directly to OpenAI's own speech-to-speech model (no orchestration platform, no per-minute markup) and are comfortable being single-vendor on OpenAI.
OpenAI's own realtime speech-to-speech models only — gpt-realtime-2.1 (flagship) or gpt-realtime-2.1-mini — no other provider is offered.
Best for Operations and CX teams at mid-market or enterprise companies who want a visual, no-code way to design and launch phone agents without an engineering ticket.
OpenAI only — GPT-4.1 mini, GPT-4.1 or GPT-5 variants depending on plan, with no other model vendor publicly documented.
Best for Organizations already running Microsoft Fabric or Power BI Premium capacity who want employees asking plain-English questions over governed lakehouse, warehouse, Power BI, mirrored and graph data without writing SQL, DAX, KQL or GQL themselves, and developers who want to embed that governed data access into a larger agent system via Azure AI Foundry, Copilot Studio, the runtime endpoint, or a service-principal-authenticated REST API.
Fixed to a GPT-4-series model behind the Azure OpenAI Assistant APIs — cannot be changed or swapped by the user.
Best for Sales reps and small-to-mid-size revenue teams that want one chat-based agent to handle prospecting, enrichment and sequencing without hiring a dedicated GTM engineer to build workflows (Clay's audience) or signing a full sales-led enterprise contract (11x's or Artisan's).
GPT-5.4 on the Free/Base/Pro tiers or GPT-5.6 on Business — always OpenAI, and not user-selectable.
Best for Professional developers who want a powerful AI agent woven into a familiar VS Code-based editor, with first-class model choice and the option to be hands-off via Cursor Router.
Claude, GPT and Gemini are all selectable per request, with no single default.
Best for Enterprise and team developers already in GitHub and mainstream IDEs who want governed, reviewable AI coding at scale, with optional access to Claude Code and OpenAI Codex through the same interface.
GPT is one of three selectable models in Copilot Chat, alongside Claude and Gemini.
Best for Engineering teams offloading well-scoped migration tasks without adding headcount
GPT runs alongside Claude, Gemini and Cognition's own SWE-tuned models under the hood.
Best for Developers who want a full, agent-native desktop IDE and the freedom to run several coding agents, including third-party ones, side by side.
GPT sits alongside Claude and Gemini as a general model option, on top of Windsurf's own SWE-tuned models, carried over into the Devin Desktop rebrand.
Best for Revenue, ops, and customer success teams running 5+ agents in production, who want multi-agent orchestration with evals, tracing, and SSO-grade governance on one bill.
GPT is one of three models its router can select per task, alongside Claude and Gemini.
Best for Researchers, analysts and knowledge workers who want fast, sourced answers, plus an autonomous research mode and an agentic browser.
GPT is one of six selectable models for Pro users, alongside Sonar, Claude, Gemini, GLM and Kimi.
Best for Salesforce service, sales, field-service and employee-support teams that need agents to act on CRM records and have Salesforce admin or developer capacity.
GPT-5.x is one of the model options admins can select, alongside Claude, Gemini or a bring-your-own-model, not the default.
Best for Teams that want a proven, enterprise-scale, compliance-certified voice AI vendor with a research-led model roadmap, and are prepared to start with a sales conversation: there is no self-serve or free way to trial it today.
GPT-5 joined PolyAI's own proprietary Raven model as a selectable option in May 2026, alongside Claude and Gemini.
Best for Engineering teams, using a coding agent or writing code by hand, that want automatic, consistent PR review plus triage, explanation and continuous security across GitHub, GitLab, Bitbucket or Azure DevOps before a human reviewer looks at it.
OpenAI models power its review backend alongside Claude Opus/Sonnet/Haiku, orchestrated internally rather than user-selectable.
Best for Large and mid-market enterprises whose biggest AI-agent blocker is fragmented, permission-sensitive knowledge scattered across 100+ internal tools, and who want employees building department-specific agents on that indexed knowledge without writing code.
Admins enable and switch to GPT from an LLM admin console, alongside Claude and Gemini, on top of Glean's own retrieval layer.
Best for A consultant, marketer, founder or operations lead who wants research, files, code, slides and media handled in one web workspace.
One of nine specialized models — with Claude, Gemini, DeepSeek and others — its Super Agent routes across and cross-checks.
Best for Support teams already committed to (or actively evaluating) Zendesk as their core helpdesk who want the AI agent to come from the same vendor and billing relationship, rather than layering a third-party agent on top.
GPT-4o is the primary backend, with Anthropic, Google and Amazon Bedrock also supported.
Best for Developers who already pay for a ChatGPT plan and want a terminal-native agent with granular, inspectable sandbox/permission controls across CLI, desktop and IDE surfaces.
OpenAI's own CLI, built on the GPT-5.6 family, with built-in support for local Ollama/LM Studio models or Amazon Bedrock as alternates.
Best for Knowledge workers who want to query data in plain language without writing SQL
GPT-5.6 is one of a short frontier set — with Claude Sonnet 5 and Claude Fable 5 — offered on top of Julius’s own models, and which of them you get depends on your plan.
Best for Ops and automation teams who want to build agentic workflows visually while keeping data on self-hosted infrastructure.
A named model-agnostic option alongside Anthropic and Ollama that its AI nodes can call.
Best for Python or .NET developers and researchers exploring multi-agent orchestration who want a mature, free, self-hosted framework.
OpenAI and Azure OpenAI are first-class in its otherwise model-agnostic `autogen-ext` design.
Best for .NET or Python teams already invested in Azure or Microsoft Foundry who want one supported agent SDK, an optional managed hosting path, and a documented migration off AutoGen or Semantic Kernel.
A named provider alongside Azure OpenAI, Anthropic Claude, Ollama and Gemini (via its Go SDK).
Best for Developers who want the fastest, lightest way to ship a straightforward agent (support triage, a tool-using assistant or a sandboxed coding agent) with tracing and guardrails included.
OpenAI's own SDK, first-class by design, extensible to 100+ other providers via LiteLLM.
Best for Python developers who want an Apache 2.0 SDK with memory, knowledge, learning, guardrails and a Control Plane they run inside their own cloud account, and who are comfortable running Docker or a cloud deploy template themselves.
One of 20+ providers reachable through its model-agnostic Python interface, alongside Anthropic, Ollama and others.
Best for Developers and teams who want an open-source, self-hostable platform to build RAG apps and agents with full data control.
One of hundreds of model-agnostic options it can connect, alongside Anthropic, Google and open/local models.
Best for Developers and AI engineers who want a free, self-hosted research agent they can embed via Python, REST or MCP and run against their own LLM and search keys.
One of several model-agnostic providers — with Anthropic, Google and local Ollama models — it can run on.
Best for Developers and product teams who want ElevenLabs' voice quality and cloning as the foundation of their agent, and are comfortable assembling the rest of the stack themselves.
One of three bring-your-own-key providers, with Anthropic and Google, or any OpenAI-compatible endpoint, or ElevenLabs' own hosted default.
Best for Developers who want hands-on visibility into what an AI coding agent is doing to their codebase and accept early-stage, self-hosted, open-source software.
One of six selectable providers — with Claude, Gemini, Ollama, OpenRouter and DeepSeek — in its visual, model-agnostic canvas.
Best for Developers who want a genuinely free, fully open-source coding agent that works with any model provider, including their own local Ollama models, without being locked into one vendor's terminal tool.
Reachable through its 75+-provider Models.dev integration, alongside Claude, Gemini and local Ollama models.
Best for Developers who want a fast, free, Go-native coding-agent harness with real session-branching control and are comfortable running fast-moving, self-described "beta forever" software from an independent maintainer.
One of 30+ named providers — with Anthropic, Gemini, Kimi, DeepSeek and more — in its self-hosted, model-agnostic CLI.
Best for Developers and teams who want one open platform that runs and coordinates multiple coding agents (Claude Code, Codex, Gemini CLI, Warp Agent) side by side across terminal, CLI and cloud, with shared team context and programmatic API access to cloud agent runs.
A named provider its terminal can orchestrate, alongside Anthropic and z.ai — and it runs OpenAI's own Codex CLI directly as a hosted agent.
Best for Developers who want a fast, open-source editor and the freedom to run Claude Code, Codex, Gemini CLI or other vendors' own agents natively via the Agent Client Protocol Zed created, instead of being locked into one editor's built-in agent.
First-class, named support alongside Anthropic, Google, Mistral, DeepSeek and more — and it natively runs OpenAI Codex as an external agent via the ACP it created.
Best for Small teams or solo operators who want pre-built AI agent roles, shared files, and OAuth-driven workflows under one credit-based subscription.
One of several selectable models — alongside Claude, Gemini, Kimi, DeepSeek or Vecbase's own — billed per token.
Best for Teams that want Google's own official, actively-developed agent SDK, especially polyglot teams (Python/Go/Java/TypeScript/Kotlin) or anyone building multi-agent systems that need to interoperate across vendors via the A2A protocol.
Reachable via LiteLLM alongside 100+ other providers, on top of native first-party Gemini support.
Best for Engineering-led teams building a multi-channel AI support agent who want an open-source, self-hostable core with an optional managed cloud layer.
One of the model-agnostic options, alongside Anthropic Claude, Google Gemini and open-weight models.
Best for Engineering teams that want a managed speech pipeline with real LLM-provider choice, and regulated enterprises that need a self-hosted or VPC deployment option a typical closed-source voice-agent vendor does not offer.
One of six bring-your-own-key LLM providers — with Anthropic, Google, NVIDIA, Groq and Amazon Bedrock — or Deepgram manages the model for you by default.
Best for Developers and teams already standardized on a JetBrains IDE who want an agent that shares the same debugger, project index, and version control already built into their editor.
Reachable via JetBrains AI credits or bring-your-own-key, alongside Anthropic and Google frontier models, plus local runtimes via Ollama/LiteLLM.
Best for Developers and self-hosters who already run their own infrastructure and want one personal AI agent that reaches WhatsApp, Telegram, Slack, Discord, Signal and iMessage instead of a separate app.
One of three hosted-account options, with Claude and DeepSeek, plus bring-your-own-key or a fully local model — genuinely model-agnostic, not GPT-first.
Best for Engineering teams building a bespoke, production-grade agent who want the largest integration ecosystem plus real stateful control.
A named model-agnostic integration alongside Anthropic and Google, documented explicitly rather than inferred.
Best for Developers and teams who want a transparent, open-source, model-agnostic agent they fully control and pay for by usage, with optional flat-rate access to open-weight models.
A named provider in its bring-your-own-key setup, alongside Anthropic, Google, OpenRouter, Bedrock, Azure, Vertex, DeepSeek, xAI, Mistral and more.
Best for Developers who want free, model-agnostic AI pair programming in their terminal
A named provider in its bring-your-own-key setup, alongside Anthropic, DeepSeek and local models.
Best for Python teams building type-safe agents who also want realtime voice, image generation and embeddings in one SDK, with optional durable execution via Temporal, DBOS, Prefect or Restate.
A named provider in its type-safe, model-agnostic interface, alongside Anthropic, Gemini, Mistral, Cohere, DeepSeek, Grok and Ollama.
Best for Open source-minded engineering teams and enterprises that want a self-hostable coding agent platform
Named by its own docs as a bring-your-own-key example alongside Claude, or use OpenHands' own hosted models on a pay-as-you-go basis.
Best for Engineering teams that want to own their voice-agent stack end to end, self-hosted or on LiveKit Cloud, without being locked into one vendor's API.
A named LLM provider in its default pipeline alongside Deepgram, Anthropic and ElevenLabs, or bring your own via LiveKit-hosted Inference — and it powers OpenAI's own ChatGPT Advanced Voice Mode under the hood.
Best for Full-stack JavaScript and TypeScript teams, especially those already building in Next.js or Node, who want agents, durable workflows and memory unified in one framework without adopting a Python-first stack.
One of the 90+ providers its model-agnostic layer supports, documented by name alongside Anthropic and Gemini.
Best for Engineering teams that want an open-source voice-agent SDK with real LLM choice, built on speech models several rival platforms in this category already license as a component.
Reachable for the Line agent layer via its LiteLLM integration (100+ providers); Cartesia's own Sonic/Ink speech models are separate and proprietary.
Best for Developers who want to stand up a role-based, collaborating multi-agent system quickly in Python, with a stable MIT-licensed core and a hosted build and runtime for production.
Model-agnostic via LiteLLM, so it can drive agents on OpenAI, Anthropic, local/open models and others, documented explicitly in its own FAQ.
Best for Devs who want an open source coding agent that reuses an existing Claude or ChatGPT plan
One of 15+ providers its ACP integration can call, alongside Anthropic, Google and Ollama, or reuse an existing ChatGPT subscription instead of a separate API key.
Best for Engineering teams building production agents who need cyclical graphs, durable state and replayable runs
OpenAI is the first provider its own docs name, alongside Anthropic and Google and an open-ended list beyond them — the same model-agnostic layer LangChain sits on.
