---
title: Models
description: Model portfolios, specifications, and published outputs.
canonical: https://productized.tech/models
lang: en
---

# Models

- [Claude Opus 5.5](https://productized.tech/models/claude-opus-5-5.md): Anthropic’s flagship successor to Opus 5, built for demanding reasoning, large-codebase coding, and long-horizon agentic work.
- [GPT-6 Sol](https://productized.tech/models/gpt-6-sol.md): The cost-efficient high-end tier of OpenAI’s GPT-6 series, between Astra and Luna, built for complex coding and agentic workflows.
- [Union Alpha](https://productized.tech/models/union-alpha.md): An anonymous preview model for research, coding, and agentic workflows, with text and image input and free access during preview.
- [Gemini 3.8 Flash](https://productized.tech/models/gemini-3-8-flash.md): Google’s high-efficiency Gemini 3 workhorse, bringing Pro-level agentic coding and multimodal execution to the Flash tier.
- [Muse Spark 1.3](https://productized.tech/models/muse-spark-1-3.md): Meta’s multimodal reasoning model for long-horizon agentic and coding workflows, with stronger first-attempt accuracy and cleaner tool use.
- [Claude Fable 5.1](https://productized.tech/models/claude-fable-5-1.md): Anthropic’s update to Fable 5, improving long-running agentic coding, knowledge work, research, and precise unattended execution.
- [DeepSeek V4 Flash Vision Exp](https://productized.tech/models/deepseek-v4-flash-vision-exp.md): DeepSeek’s experimental vision add-on to V4 Flash 0731, adding image understanding while matching the base model on text, agents, and reasoning.
- [Ox Alpha](https://productized.tech/models/ox-alpha.md): A stealth reasoning model on OpenRouter, built for coding and long-horizon agent work, currently free while the vendor stays unnamed.
- [GLM-5.3](https://productized.tech/models/glm-5-3.md): Z.ai’s latest flagship, post-trained from the GLM-5.2 base for complex coding and long-horizon agent work.
- [Gemini 3.7 Flash](https://productized.tech/models/gemini-3-7-flash.md): Google’s latest Flash workhorse, shipped three weeks after 3.6 with stronger coding and agent performance.
- [DeepSeek V4 Pro](https://productized.tech/models/deepseek-v4-pro-0813.md): DeepSeek’s flagship V4 model, now generally available with stronger agent performance as the 0813 checkpoint.
- [Grok 4.6](https://productized.tech/models/grok-4-6.md): xAI’s follow-up to Grok 4.5, tuned for long-running agents and more ambitious interactive and visual work.
- [DeepSeek V4 Flash](https://productized.tech/models/deepseek-v4-flash-0731.md): DeepSeek’s cost-efficient V4 workhorse, re-post-trained for agents as the official 0731 checkpoint.
- [Gemini 3.6 Flash](https://productized.tech/models/gemini-3-6-flash.md): Google’s more token-efficient successor to 3.5 Flash, built for coding, knowledge work, and multimodal agents.
- [Kimi K3](https://productized.tech/models/kimi-k3.md): Moonshot AI’s 2.8T-parameter flagship — native vision, a 1M-token context, and the first open 3T-class model.
- [GPT-5.6 Sol Pro](https://productized.tech/models/gpt-5-6-sol-pro.md): OpenAI’s flagship GPT-5.6 Sol in pro reasoning mode — the highest-capability option for difficult, longer-running work.
- [Grok 4.5](https://productized.tech/models/grok-4-5.md): xAI’s coding- and agent-focused frontier model that made xAI a serious contender in the AI space.
- [Gemini 3.5 Flash](https://productized.tech/models/gemini-3-5-flash.md): Google’s I/O 2026 Flash model — frontier agent and coding performance at Flash speed, with a 1M-token context.

Use find_models and compare_models on the HTML Models page to filter models and open comparisons.