Models

OpenAI: GPT-6 Astra Pro

NewT

55.3M tokens

OpenAI's most capable frontier reasoning and multimodal model for highly complex workflows, software engineering, and scientific tasks.

OpenAI: GPT-6 Astra

NewT

1.1M tokens

GPT-6 Astra is OpenAI's flagship model for demanding end-to-end work. It is suited for advanced analysis, software engineering, deep research, and long-horizon tasks.

OpenAI: GPT-5.6 Sol

T

18.2M tokens

Frontier balance between deep reasoning, extensive context, and real-time generation speed for everyday production workflows.

OpenAI: GPT-5.6 Terra

T

32.8M tokens

High-efficiency frontier model optimized for enterprise data pipelines, document extraction, and long-form analytical synthesis.

OpenAI: GPT-5.6 Luna

50% offT

89.4M tokens

Ultra-fast low-cost model with full 1M+ context window for high-volume automation and classification.

OpenAI: GPT-5

T

24.1M tokens

Powerful reasoning and general intelligence engine with multimodal vision support for creative and analytical generation.

OpenAI: GPT-5 mini

T

142.0M tokens

Cost-effective, low-latency reasoning and conversational intelligence for consumer-facing apps.

OpenAI: GPT-5 nano

T

61.3M tokens

Ultra-compact sub-second execution model for edge tasks, summarization, and content classification.

OpenAI: GPT-4.1

T

74.5M tokens

Reliable workhorse model with massive 1M token context for long document synthesis and legal review.

OpenAI: GPT-4.1 mini

T

115.8M tokens

Affordable 1M-context model for processing large codebases and complex multi-document repositories.

OpenAI: GPT-4.1 nano

T

44.1M tokens

Lowest latency 1M-context variant for lightweight summarization and routing pipelines.

OpenAI: GPT-4o

T

310.2M tokens

Flagship omni-channel model for reasoning across text, code, and vision with low latency.

OpenAI: GPT-4o mini

T

580.4M tokens

High-speed, cost-efficient model for lightweight multimodal operations and structured outputs.

OpenAI: o3

T

98.7M tokens

Advanced deliberate reasoning model tailored for science, math, and competitive programming benchmarks.

OpenAI: o4-mini

T

82.4M tokens

Fast and affordable deep reasoning model designed for STEM, theorem proving, and analytical logic.

OpenAI: GPT OSS 120B

T

38.6M tokens

Open-weights foundation model offering transparency, fine-tunability, and rapid throughput.

Anthropic: Claude Opus 4.7

T

76.4M tokens

Anthropic's pinnacle intelligence for deep analysis, creative synthesis, and multi-step complex coding.

Anthropic: Claude Sonnet 4

T

245.9M tokens

Top-tier balance of frontier intelligence, speed, and vision reasoning for daily software engineering.

Anthropic: Claude 3.5 Sonnet v2

T

190.1M tokens

Industry-leading model for computer use, code generation, and visual understanding.

Amazon: Amazon Nova Premier

T

42.0M tokens

Amazon's most capable multimodal model for complex reasoning across text, image, and high-framerate video.

Amazon: Amazon Nova Pro

T

58.1M tokens

High-accuracy multimodal model tailored for enterprise visual reasoning, documents, and agentic tasks.

Amazon: Amazon Nova Lite

50% offT

128.5M tokens

Very low-cost, lightning-fast multimodal processing across text, image, and video media streams.

Amazon: Amazon Nova Micro

T

210.0M tokens

Text-only model offering lowest latency and rock-bottom inference costs for classification and intent routing.

Google's high-efficiency open model engineered for versatile linguistic and code understanding.

Ultra-compact open model optimized for lightweight tasks and responsive chat completions.

Open-architecture reasoning model matching top frontier models on math, code, and logic benchmarks.

Next-iteration general LLM with enhanced instruction following and agentic workflows.

High-performance Mixture-of-Experts architecture delivering frontier output at extreme value.

Meta's flagship open multimodal model with 1M context window and vision comprehension.

Proven open-weights model delivering industry benchmark performance in coding and reasoning.

Frontier model with extensive world knowledge, sharp reasoning, and unfiltered analytical depth.

Huge 2M token context window paired with lightning-fast token generation speed.

Mistral AI: Mistral Large 3

T

87.3M tokens

Mistral's flagship multilingual model with state-of-the-art reasoning and native function calling.

Mistral AI: Devstral 2 123B

T

104.9M tokens

Engineered specifically for complex software architecture, multi-file refactoring, and code review.

Mistral AI: Mistral Medium 3.5

T

52.4M tokens

Well-rounded performance across European languages, logic reasoning, and enterprise formatting.

Microsoft: Phi-4

T

41.9M tokens

High-density small language model with state-of-the-art synthetic-data reasoning abilities.

Microsoft: Phi-4 mini

T

63.7M tokens

Lightweight SLM engineered for rapid text processing, categorization, and routing.

Microsoft: Phi-4 multimodal

T

38.2M tokens

Compact multi-modal model handling simultaneous text, visual diagrams, and speech audio.

Microsoft: MAI-DS-R1

T

28.4M tokens

Microsoft-tuned DeepSeek R1 reasoning distribution for enterprise analytical tasks.

Specialized long-context model capable of lossless document search and summarization.

Deliberate chain-of-thought architecture for intricate logic, research, and coding tasks.

Fluid conversational and creative writing engine with high token generation throughput.

NVIDIA-accelerated LLM built for high-throughput enterprise synthesis and data curation.

Next-generation sparse MoE architecture with expansive multilingual understanding.

Advanced visual-language model adept at document OCR, charts, and spatial understanding.

Bilingual powerhouse for Chinese and English reasoning, translation, and code synthesis.

High-speed bilingual conversational model for production latency requirements.

Enterprise-optimized LLM specifically tuned for Retrieval-Augmented Generation (RAG) and tool use.