OpenAI

GPT-6.1 Sol approaches GPT-6 Astra on demanding evals at Sol pricing. Hybrid reasoning with five effort levels (low, medium, high, xhigh, max), mid-conversation effort and task-budget controls, efficient code generation, agentic search, computer use, tool use, text and image inputs, steering, memory, and subagent coordination.

GPT-6 Astra is OpenAI's most capable model, built for the hardest end-to-end work: complex reasoning, coding, computer use, research, and document creation. Five reasoning effort levels (low, medium, high, xhigh, max), text and image inputs, tools, and a 1.05M-token context window.

GPT-6 Sol supports complex coding and multi-step workflows across tools and applications. Text and image inputs, text outputs including code, and adjustable reasoning effort.

GPT-6 Luna is the cost-efficient GPT-6 tier for focused, high-volume work. Text and image inputs, text outputs including code, and adjustable reasoning effort.

GPT-5.6 Sol is OpenAI's flagship frontier model in the GPT-5.6 family, built for faster, more reliable results on complex professional work. Text and image inputs, 1M context, reasoning, server-side tool calling, and prompt caching.

GPT-5.6 Terra is the balanced GPT-5.6 tier for everyday production work. Text and image inputs, 1M context, reasoning, server-side tool calling, and prompt caching.

GPT-5.6 Luna is the fast, low-cost GPT-5.6 model for high-volume work. Text and image inputs, 1M context, reasoning, server-side tool calling, and prompt caching.

Maximum capability variant of GPT-5.5 for research-grade and highest-stakes tasks. No cached-input discount available. Data residency endpoints incur a 10% price uplift.

GPT-5.5 is OpenAI's model for complex professional work: advanced coding, research, analysis, software operation, document workflows, and long-running agentic tasks. Text and image input, text output, 1,050,000-token context, streamed responses, and function calling.

Premium GPT-5.4 Pro for hardest problems — higher compute, no cached-input discount.

GPT-5.4 brings frontier reasoning, coding, computer use, long-context workflows, and tool use to professional workflows. Text and image input, text output, streamed responses, and function calling.

Ultra-efficient GPT-5.4 nano tier for high-volume workloads with vision and tools.

Compact GPT-5.4 mini for coding, agentic subtasks, and multimodal chat with long context.

Efficient GPT-5 family model on Azure OpenAI (Global Standard). Balanced reasoning, vision, and tool use at lower cost than GPT-5.

Flagship GPT-5 model — general-purpose frontier reasoning with vision, tools, and long context.

Cheapest model in OpenAI's lineup. $0.10/MTok input is competitive with the cheapest models on the market. Ideal for routing, extraction, and lightweight classification. Remains competitive even after GPT-5.5 launch.

Cost-efficient multimodal GPT-4o mini with vision and tools. Azure OpenAI Global Standard.

Compact o-series reasoning model on Azure OpenAI (Global Standard). Excels at math, coding, and visual reasoning with a 200K context window.

Maximum reasoning capability in OpenAI's lineup. For workloads where correctness is paramount over cost. No cached-input or batch discount available.

o-series reasoning model for complex math, science, coding, and visual reasoning. Succeeded by GPT-5 for many workloads.

Lightweight reasoning model. Positioned between o4-mini and o3 for cost/capability trade-offs. Context window expanded to 200K tokens.

Most cost-effective option for 1M-token context workloads. Best choice for long-document processing when cost matters. Reliable production model.

Best value for long-context tasks at low cost. 1M context window at only $0.40/MTok input.

Multimodal GPT-4o chat model with vision, tools, and 128K context. Azure OpenAI Global Standard.

Flagship OpenAI image generation model. Best quality in the GPT Image 1.x lineup; text-to-image, editing, Low/Medium/High quality tiers.

Cost-efficient OpenAI image generation model. Text-to-image and editing via Images API; token-priced text and image modalities.

Ultra-low-cost GPT-5 family model for high-volume lightweight tasks with vision and tools.

Premium GPT-5 Pro tier for hardest problems — higher compute, no cached-input discount on standard pricing.

GPT-5.3 Codex agentic coding model on Azure OpenAI Global Standard (Responses API).

Most capable OpenAI embedding model for English and multilingual semantic search, clustering, recommendations, and classification. Azure OpenAI Global Standard deployment.

Efficient OpenAI embedding model — improved successor to text-embedding-ada-002. Azure OpenAI Global Standard deployment.

Latest OpenAI image generation model. Text-to-image and editing with separate text/image token pricing. Azure Global Standard.

ChatGPT alias for the latest GPT Image model in OpenAI APIs (Standard). Same token pricing as GPT Image 1.5 for text and image modalities.

OpenAI live speech-to-text for realtime transcription sessions. Billed per minute of audio.

High-accuracy OpenAI speech-to-text for file transcription and committed realtime turns. Supports keyword/language hints.

Lower-cost OpenAI speech-to-text for batch file transcription via POST /v1/audio/transcriptions.

Higher-quality OpenAI speech-to-text for file transcription via POST /v1/audio/transcriptions.

OpenAI steerable text-to-speech model (token-priced) via POST /v1/audio/speech.

Cost-efficient GPT Audio for Chat Completions — accepts text/audio inputs and returns text/audio outputs.

Cost-efficient realtime voice model over WebRTC/WebSocket/SIP — text, audio, and image inputs with realtime audio/text outputs.

GPT-5 optimized for agentic coding in Codex / Responses API environments.

GPT-4o speech-to-text with speaker diarization for multi-speaker audio files.

Smaller, cost-efficient GPT-5.1 Codex for agentic coding via the Responses API.

OpenAI video generation model (text/image to video). Billed per second of generated video at 720p.

GPT-5.1 optimized for agentic coding in Codex / Responses API.

GPT-5.1 flagship for coding and agentic tasks with configurable reasoning effort (none to high).

GPT-5.1 Codex variant optimized for long-running agentic coding tasks via the Responses API.

GPT-5.2 frontier chat/reasoning model with vision and tools. Azure OpenAI Global Standard.

GPT-5.2 Codex optimized for agentic coding via the Responses API. Azure Global Standard.

OpenAI GPT Audio 1.5 for Chat Completions with text/audio inputs and text/audio outputs. Azure Global Standard.

OpenAI realtime voice model 2.1 over WebRTC/WebSocket/SIP with text, audio, and image inputs.

Cost-efficient realtime voice model 2.1 mini with text, audio, and image inputs.

GPT OSS Safeguard 20B is OpenAI's compact open-weight safety model for lightweight content moderation, policy reasoning, content filtering, risk analysis, and justified guidance.

GPT OSS Safeguard 120B is OpenAI's 120B open-weight safety model for advanced safety reasoning, nuanced policy interpretation, multi-turn safety analysis, and justified decisions.

gpt-oss-120b is OpenAI's 120B open-weight model for text generation, coding, and reasoning. Hybrid reasoning, extended thinking, tool use, and agentic search.

gpt-oss-20b is OpenAI's 20B open-weight model for efficient text generation, coding, and reasoning at lower cost.