# Frontier AI models: May 2026

> No new flagship shipped in May - the frontier held at Claude Opus 4.7, GPT-5.5 and Gemini 3.1 Pro while the industry absorbed April.

## Direct answer

No frontier AI model was released in May 2026. The current frontier models through May were Claude Opus 4.7 (Anthropic), GPT-5.5 (OpenAI), Gemini 3.1 Pro (Google) and DeepSeek V4 (open source), all shipped in the preceding weeks. May was a consolidation month: the labs were quiet and adoption caught up with the April Sprint. This is the normal shape of the cycle - a dense release cluster followed by a month of deployment, not a slowdown.

## Released in May 2026

No frontier model was released in May 2026.

## Leading models as of May 2026

| Lab | Leading model | Strongest at |
| --- | --- | --- |
| Anthropic | Claude Opus 4.7 | Sustained multi-file software engineering |
| OpenAI | GPT-5.5 | Autonomous multi-step research and agentic workflows |
| Google | Gemini 3.1 Pro | Multimodal reasoning and throughput |
| DeepSeek | DeepSeek V4 | Near-frontier capability at open-source cost |

## What this month changed for careers

A quiet month at the labs is not a quiet month at work. May is when the April models actually arrived in the tools people use, and adoption - not capability - is what changes anyone's job. The gap between 'a model can do this' and 'my employer has deployed a model that does this' is measured in months like May.

## Questions

### What frontier AI models were current in May 2026?

Claude Opus 4.7 from Anthropic, GPT-5.5 from OpenAI, Gemini 3.1 Pro from Google and the open-source DeepSeek V4. All four shipped between February and April 2026 and remained the leading models through May.

### Was any AI model released in May 2026?

No frontier model launched in May 2026. It followed the April Sprint, in which seven models shipped in one month, and served as a consolidation period while deployment caught up.

### What was the best AI model in May 2026?

It depended on the task. Claude Opus 4.7 led software engineering, GPT-5.5 led multi-step research and agentic work, and Gemini 3.1 Pro led multimodal reasoning. No single model led across all categories.

---

**Source:** Job Security Meter — https://jobsecuritymeter.com/ai-model-releases/may-2026
**Human-readable version:** https://jobsecuritymeter.com/ai-model-releases/may-2026
**Last updated:** May 2026
**Attribution:** Free to quote and cite with attribution to Job Security Meter.
**Full site specification for language models:** https://jobsecuritymeter.com/llms-full.txt