AI Timeline - 194+ LLMs on One Interactive Chart

TL;DR

TL;DR: AI Timeline is a free, interactive chart that maps 194+ large language models chronologically from the original Transformer (2017) to GPT-5.3 (2026), making it easy to see how the field evolved.

What Is AI Timeline?

AI Timeline is a free interactive visualization that tracks every major large language model released since 2017. It maps 194+ models chronologically, showing their release dates, parameter counts, context windows, and how they relate to each other across the AI landscape.

The site describes itself as:

“Interactive timeline tracking 194+ Large Language Models from 2017 to 2026, including ChatGPT, GPT-4, Claude, Gemini, LLaMA, Mistral, and DeepSeek.”

The timeline starts with the original Transformer paper (2017) and ends with GPT-5.3 (2026), giving a clear visual narrative of how the field exploded from a research paper into a global industry.

Source and Accuracy Notes

⚠️ This section is MANDATORY. All links must be verified from actual source, not guessed.

• Project page: llm-timeline.com — verified live
• Source repository: none (web-only project)
• License: not applicable — informational reference site
• HN launch thread: news.ycombinator.com/item?id=41892593 — 174 points, verified via HN Algolia API

How to Use the Timeline

The interface is straightforward:

1. Open llm-timeline.com in a browser
2. Scroll horizontally through the chronological timeline
3. Click any model node to see details: release date, developer, parameter count (if known), context window, and a link to the model page
4. Use the zoom controls to focus on a specific era (e.g. 2022-2024)

What Models Are Covered

The timeline spans every major release across the AI industry:

• Google DeepMind: LaMDA, PaLM, Gemini 1.0 through 2.0
• OpenAI: GPT-2, GPT-3, ChatGPT (GPT-3.5), GPT-4, GPT-4o, o-series models, GPT-5.3
• Anthropic: Claude 1 through Claude 4, including Opus, Sonnet, and Haiku variants
• Meta: LLaMA, LLaMA 2, LLaMA 3 series
• Mistral AI: Mistral, Mixtral, Mistral Large, Mistral Small
• DeepSeek: DeepSeek Coder, DeepSeek V2, V3, R1
• Open-source models: Falcon, Vicuna, Orca, Phi series, and dozens more

The count grew from 171 models at launch to 194+ as new releases came out.

Key Historical Milestones

The timeline makes it easy to see how rapidly the field moved:

• 2017: Transformer paper launches the foundation
• 2018: BERT and GPT-1 establish the two dominant architectures
• 2019: GPT-2 raises concerns about open release
• 2020: GPT-3 demonstrates emergent capabilities at scale
• 2022: ChatGPT (GPT-3.5) brings AI to mainstream attention
• 2023: GPT-4, Claude 2, Gemini 1.0, open-source Llama models
• 2024: GPT-4o, Claude 3, Gemini 1.5, DeepSeek V2, Llama 3
• 2025-2026: o-series reasoning models, GPT-5.3, Gemini 2.0, Claude 4

Why the Timeline Is Useful

There is no single canonical reference for all LLM releases. Models are announced via blog posts, papers, API updates, and news articles — scattered across different sources. AI Timeline consolidates them into one searchable, visual reference.

This is useful for:

• Researchers tracking the evolution of model capabilities and architectures
• Developers choosing which models to integrate based on release chronology and specs
• Writers and journalists needing accurate release timelines for articles
• Anyone curious about how we got from the 2017 Transformer to GPT-5.3 in under a decade

FAQ

Q: Is the timeline complete? A: No. It tracks major releases from established labs and notable open-source projects. Smaller fine-tunes, region-specific models, and enterprise-only releases are not included. The 194+ count is accurate as of September 2026.

Q: Are parameter counts verified? A: Parameter counts are included where they have been officially confirmed or reliably estimated. Many labs (notably Anthropic and Google DeepMind) do not publish parameter counts, so those entries are marked as unknown.

Q: Can I download the data? A: The site does not offer a data export. The underlying data appears to be maintained in a private dataset.

Q: Is this affiliated with any AI lab? A: No. AI Timeline is an independent project by a developer, as noted in the HN launch thread.

Conclusion

AI Timeline fills a real gap — there is no official registry of LLM releases, and tracking the field by memory leads to errors. Having a chronological view from the original Transformer through GPT-5.3 makes it easy to understand the pace of development and how different labs influenced each other.

Bookmark llm-timeline.com as a reference for your next article, presentation, or research project.