AI Models Directory

OpenAI GPT-6 model for focused, high-volume workloads where cost efficiency is the priority.

Anthropic Claude model for long-running agentic coding and knowledge work.

OpenAI GPT-6 model built for complex coding and agentic workflows, balancing intelligence and cost.

The latest voice foundation model from Google DeepMind, integrating extended chain-of-thought reasoning directly into live spoken dialogues.

TypeSafe current System One model for fast, structured decisions with typed outputs, probabilities, and confidence.

DeepSeek current native-multimodal Flash model with an asymmetric 552B-parameter MoE architecture.

The latest evolution in the Gemini 3 family, delivering state-of-the-art software engineering (73.7% DeepSWE) and agentic enterprise knowledge workflows.

The ultimate model for demanding reasoning and long-horizon agentic work, featuring 0.025x cache reads ($0.25/MTok).

The most capable autonomous intelligence system engineered by Anthropic, restricted to Project Glasswing participants.

Universal omnimodal generation model that accepts any combination of text, audio, image, and video to generate any combination of outputs.

Ultra-low latency version of GLM-5.3-Flash engineered for high-concurrency real-time IDE code completion and interactive agents.

The first native multimodal model in the GLM-5 series: 320B parameters (18B active) combining linear and sparse attention for low-cost visual coding.

Universal multilingual speech engine supporting live simultaneous translation and speaker diarization across 100+ languages.

OpenAI’s premier sixth-generation frontier model, built for the most demanding end-to-end software engineering and scientific research tasks.

Z.AI’s premier flagship model, delivering a 50% performance gain on Code Bench and matching Claude Mythos 5 in cybersecurity and vulnerability discovery.

High-performance multimodal foundation model featuring algorithmic reasoning enhancements and agentic long-form video understanding.

DeepSeek V4 Pro API model, version DeepSeek-V4-Pro-0813, with thinking and non-thinking modes.

Cloud embodied foundation model capable of synthesizing novel robotic tool use and executing complex mechanical assemblies.

Second-generation on-device physical AI foundation model incorporating tactile sensor feedback for delicate robotic manipulation.

The premier enterprise flagship model for multi-hour autonomous coding, complex system engineering, and vision-heavy agent tasks.

Iterative performance leap in the Flash family featuring major upgrades in code refactoring and autonomous agent tool loops.

High-speed, high-density lightweight model providing strong analytical accuracy for 8 cents per million tokens.

Moonshot AI’s premier flagship model with 2.8 trillion parameters, Kimi Delta Attention, 1M context, and open weights.

Sub-second, sub-cent visual generation model designed for high-scale gaming asset generation, thumbnails, and preview workflows.

Anthropic’s fifth-generation workhorse, combining 1M context with adaptive thinking on by default at a lower $2/$10 price point.

Flagship model family of mid-2026 featuring Sol (flagship), Terra (balanced), Luna (high-speed), and Cyber (authorized vulnerability research).

Flagship model built for project-scale context, supporting truly usable 1M-token context with 128k output.

High-speed variant of Kimi K2.7 Code with generation speeds of 180 to 260 tokens per second for instantaneous coding feedback.

Dedicated coding model delivering higher task success rates, tighter instruction following, and a 30% reduction in overthinking.

Fifth-generation Claude architecture engineered specifically for multi-hour autonomous agent workflows and deep research.

Limited-availability autonomous intelligence model delivering breakthrough problem-solving across scientific and technical disciplines.

The pinnacle of the Claude 4 generation, serving as the universal recommended replacement for legacy Claude 2 and Claude 3 Opus.

Mid-generation multimodal speed model with configurable thinking levels, allowing developers to balance latency and reasoning depth.

OpenAI’s premier full-duplex voice model for natural, expressive spoken dialogues with instantaneous interruption handling.

Extended Reasoning (ER) foundation model enabling fine-motor tool use and dynamic physics adaptation for autonomous robotic arms.

A new class of intelligence for advanced coding and professional knowledge work with a 1.05M context window and 128K output capacity.

The first Opus model with 1,000,000 tokens of context and 128,000 tokens of maximum output generation.

Dedicated conversational voice model providing native bidirectional speech streaming with 220ms end-to-end latency.

General-purpose model supporting text, image, and video inputs, switchable thinking modes, and autonomous agent tasks over 256k context.

Engineered for long-horizon tasks, able to work independently for up to 8 hours in a single run, aligned with Claude Opus 4.6.

OpenAI’s premier image synthesis and conversational in-painting model, supporting pinpoint region edits and variable quality modes.

A preview of Anthropic’s next-generation autonomous reasoning system with 1M context, deployed for frontier safety evaluation.

First OpenAI model family introducing a 1.05-million-token context window with configurable reasoning effort and affordable pricing.

The Gemini 3.1 family’s budget workhorse, delivering solid reasoning and vision parsing at 6 cents per million tokens.

High-speed image generation and editing model engineered for conversational chats and real-time interactive design.

Google’s most advanced model for complex coding, scientific discovery, and long-horizon multimodal reasoning with a 64K output token buffer.

The designated LTS upgrade target for all Claude 3 and early Claude 4 applications, offering 500k context.

Fifth-generation foundation model shifting from coding to complex systems engineering, benchmarked against Claude Opus 4.5.

Refined reasoning engine delivering over 80% on SWE-bench Verified with advanced tool choice control.

Compact optical character recognition model combining CogViT with GLM-0.5B for fast, highly accurate document and table extraction.

Free-tier version of GLM-4.7 delivering fast inference, 200k context, and strong coding for high-frequency applications.

State-of-the-art image generation model combining autoregressive semantic understanding with diffusion decoding for accurate text rendering.

Optimized agentic coding model setting open-source SOTA performance on major coding and reasoning benchmarks with 200k context.

The third-generation speed-optimized foundation model combining sub-second latency with near-Pro intelligence.

Multimodal mobile automation framework that understands smartphone screen content and executes real user actions through ADB across 50+ mainstream apps.

High-accuracy automatic speech recognition model delivering a Character Error Rate as low as 0.0717 with user-defined vocabulary support.

Multimodal vision model with native function calling and controllable thinking mode switch over 128k tokens.

Professional-grade visual synthesis model featuring flawless text rendering, spatial consistency, and multi-turn iterative image editing.

The inaugural model in the Gemini 3 family, establishing a new frontier in native multimodal reasoning and autonomous coding.

A major architectural refinement reducing Opus pricing by 66% while increasing context to 500,000 tokens.