is the next model out?

Audience measurement

Visits are counted for the public audience. You can opt out.

moreless

Page views, country, device type and browser. No advertising cookie, no cross-site tracking, no third-party tool. A random local ID stays 13 months, so you are counted once. Records are kept 25 months. You can stop everything here.

Every switch below is yours. Aligned with GDPR and the EU ePrivacy rules, the CCPA for the United States and privacy laws worldwide. Read the page.

Is the next AI model out yet? Live release tracker

no, not yet

still waiting… but the cats believe

GPT

Next model

GPT-6.1 Astra

rumour · Estimate
EstimateNov / Dec

Release alerts, no account.

Latest release

GPT-6.1 Terra · 8 days ago

This week

Have you used it? GPT-6.1 Terra

The next token seems to be “soon”.

According to @synthwavedd

8 sources checked · 3 min ago

Claude

Next model

Claude Fable 5.5

rumour · Estimate
EstimateNov / Dec

Release alerts, no account.

Latest release

Claude Haiku 5.5 · 27 days ago

This week

Have you used it? Claude Haiku 5.5

Fable is still writing its last chapter.

According to @Mr_Salio

6 sources checked · 3 min ago

Grok

Next model

Grok 4.8

rumour · 33 days late
Deadline passedOct 1, 2026
33d02h38m43s

Release alerts, no account.

Latest release

Grok 4.7 · 43 days ago

This week

Have you used it? Grok 4.7

Grok is still negotiating with the clock.

According to cellcog.ai

3 sources checked · 4 min ago

DeepSeek

Next model

DeepSeek V4.1 Pro

no date

Release alerts, no account.

Latest release

DeepSeek V4.1 Flash 1017 · 17 days ago

This week

Have you used it? DeepSeek V4.1 Flash 1017

The whale is gathering momentum.

According to cellcog.ai

4 sources checked · 3 min ago

Qwen

Next model

Qwen 4

rumour · late
Deadline passedSep 22, 2026
42d02h38m43s

Release alerts, no account.

Latest release

Qwen3.8 Omni Flash · 46 days ago

This week

Have you used it? Qwen3.8 Omni Flash

One more epoch before the big reveal?

According to CellCog

6 sources checked · 3 min ago

GLM

Next model

GLM-5.5

rumour · no day

Release alerts, no account.

Latest release

GLM 5.3 Flash · 69 days ago

This week

Have you used it? GLM 5.3 Flash

The suspense is already available.

According to @kimmonismus

4 sources checked · 3 min ago

Model history

Choose a lab to explore its releases.

GPTOpenAI

33 releases · last 8 days ago · 41.8 d avg · 335 d longest

  1. today Nov 3, 2026

  2. GPT-6.1 Terra

    out

    • context 1.05M tokensContext window of GPT-6.1 Terra. GPT-6.1 Sol is the previous model of the same tier, but the gap is beyond what a percentage explains.
    • price -50 %Prompt price compared with GPT-6.1 Sol, same tier in this family. Went from $2.00 / MTok to $1.00 / MTok. Lower is better.

    GPT-6.1 Terra is the mid-tier model in OpenAI's GPT-6.1 series, priced below GPT-6.1 Sol. It is suited for everyday agentic, coding, and writing work, is particularly strong at understanding screenshots...

    @OpenAI · X

  3. GPT-6.1 Sol

    out

    • context 1.05M tokensContext window of GPT-6.1 Sol. GPT-6 Sol is the previous model of the same tier, but the gap is beyond what a percentage explains.
    • price $2.00 / MTokPrompt price of GPT-6.1 Sol. GPT-6 Sol is the previous model of the same tier, but the gap is beyond what a percentage explains.

    GPT-6.1 Sol is an upgrade to GPT-6 Sol from OpenAI, positioned below the flagship GPT-6 Astra in the GPT-6 series. It is suited for agentic coding, computer use, document-heavy professional...

    OpenAI · GPT-6.1 Sol

  4. 2 models on this date

    GPT-6 Sol

    out

    • context 1.05M tokensContext window of GPT-6 Sol. GPT-6 Astra is the previous model of the same tier, but the gap is beyond what a percentage explains.
    • price -80 %Prompt price compared with GPT-6 Astra, same tier in this family. Went from $10.00 / MTok to $2.00 / MTok. Lower is better.

    GPT-6 Sol is the cost-efficient high-end model in OpenAI's GPT-6 series, positioned below the flagship GPT-6 Astra and above the fast GPT-6 Luna tier. It is suited for demanding professional...

    OpenAI · GPT-6 Sol & Luna

    GPT-6 Luna

    out

    • context 1.05M tokensContext window of GPT-6 Luna. GPT-6 Astra is the previous model of the same tier, but the gap is beyond what a percentage explains.
    • price -99 %Prompt price compared with GPT-6 Astra, same tier in this family. Went from $10.00 / MTok to $0.100 / MTok. Lower is better.

    GPT-6 Luna is the fast, cost-efficient model in OpenAI's GPT-6 series, positioned below GPT-6 Sol. It is suited for high-volume and latency-sensitive workloads such as chat, classification, and lightweight agentic...

    OpenAI · GPT-6 Sol & Luna

  5. GPT-6 Astra

    out

    • context 1.05M tokensContext window of GPT-6 Astra. GPT-5.6 Luna is the previous model of the same tier, but the gap is beyond what a percentage explains.
    • price $10.00 / MTokPrompt price of GPT-6 Astra. GPT-5.6 Luna is the previous model of the same tier, but the gap is beyond what a percentage explains.

    GPT-6 Astra is OpenAI's flagship model for demanding end-to-end work. It is suited for advanced analysis, software engineering, deep research, scientific work, and document creation, with particular strengths in long-horizon...

    OpenAI · GPT-6 Astra

  6. 3 models on this date

    GPT-5.6 Luna

    out

    • context 1.05M tokensContext window of GPT-5.6 Luna. GPT-5.5 is the previous model of the same tier, but the gap is beyond what a percentage explains.
    • price -96 %Prompt price compared with GPT-5.5, same tier in this family. Went from $5.00 / MTok to $0.200 / MTok. Lower is better.

    GPT-5.6 Luna is a fast, cost-efficient model in OpenAI's GPT-5.6 series. It is suited for high-volume, latency-sensitive tasks such as chat, classification, and lightweight agentic workflows, providing capable reasoning for...

    OpenAI · GPT-5.6

    GPT-5.6 Terra

    out

    • context 1.05M tokensContext window of GPT-5.6 Terra. GPT-5.5 is the previous model of the same tier, but the gap is beyond what a percentage explains.
    • price -60 %Prompt price compared with GPT-5.5, same tier in this family. Went from $5.00 / MTok to $2.00 / MTok. Lower is better.

    GPT-5.6 Terra is a balanced model in OpenAI's GPT-5.6 series, positioned between the flagship Sol tier and the cost-efficient Luna tier. It is suited for everyday coding, reasoning, and agentic...

    OpenAI · GPT-5.6

    GPT-5.6 Sol

    out

    • context 1.05M tokensContext window of GPT-5.6 Sol. GPT-5.5 is the previous model of the same tier, but the gap is beyond what a percentage explains.
    • price -20 %Prompt price compared with GPT-5.5, same tier in this family. Went from $5.00 / MTok to $4.00 / MTok. Lower is better.

    GPT-5.6 Sol is the flagship model in OpenAI's GPT-5.6 series. It is suited for complex reasoning, coding, and agentic workflows, and is particularly strong at command-line and multi-step coding tasks...

    OpenAI · GPT-5.6

  7. GPT-5.5 Pro

    catalogued

    Release date unconfirmed

    • context 1.05M tokensContext window of GPT-5.5 Pro. GPT-5.4 Pro is the previous model of the same tier, but the gap is beyond what a percentage explains.
    • price $30.00 / MTokPrompt price of GPT-5.5 Pro. GPT-5.4 Pro is the previous model of the same tier, but the gap is beyond what a percentage explains.

    GPT-5.5 Pro is OpenAI’s high-capability model optimized for deep reasoning and accuracy on complex, high-stakes workloads. It features a 1M+ token context window (922K input, 128K output) with support for...

    OpenRouter catalogue

  8. GPT-5.5

    out

    • context 1.05M tokensContext window of GPT-5.5. GPT-5.4 is the previous model of the same tier, but the gap is beyond what a percentage explains.
    • price +100 %Prompt price compared with GPT-5.4, same tier in this family. Went from $2.50 / MTok to $5.00 / MTok. Lower is better.

    GPT-5.5 is OpenAI’s frontier model designed for complex professional workloads, building on GPT-5.4 with stronger reasoning, higher reliability, and improved token efficiency on hard tasks. It features a 1M+ token...

    OpenAI · GPT-5.5

  9. 2 models on this date

    GPT-5.4 Nano

    out

    • context 400k tokensContext window of GPT-5.4 Nano. GPT-5 Nano is the previous model of the same tier, but the gap is beyond what a percentage explains.
    • price +300 %Prompt price compared with GPT-5 Nano, same tier in this family. Went from $0.050 / MTok to $0.200 / MTok. Lower is better.

    GPT-5.4 nano is the most lightweight and cost-efficient variant of the GPT-5.4 family, optimized for speed-critical and high-volume tasks. It supports text and image inputs and is designed for low-latency...

    OpenAI · GPT-5.4 mini and nano

    GPT-5.4 Mini

    out

    • context 400k tokensContext window of GPT-5.4 Mini. GPT-5.1-Codex-Mini is the previous model of the same tier, but the gap is beyond what a percentage explains.
    • price +200 %Prompt price compared with GPT-5.1-Codex-Mini, same tier in this family. Went from $0.250 / MTok to $0.750 / MTok. Lower is better.

    GPT-5.4 mini brings the core capabilities of GPT-5.4 to a faster, more efficient model optimized for high-throughput workloads. It supports text and image inputs with strong performance across reasoning, coding,...

    OpenAI · GPT-5.4 mini and nano

  10. GPT-5.4

    out

    • context +163 %Context window compared with GPT-5.3-Codex, same tier in this family. Went from 400k tokens to 1.05M tokens.
    • price +43 %Prompt price compared with GPT-5.3-Codex, same tier in this family. Went from $1.75 / MTok to $2.50 / MTok. Lower is better.

    GPT-5.4 is OpenAI’s latest frontier model, unifying the Codex and GPT lines into a single system. It features a 1M+ token context window (922K input, 128K output) with support for...

    OpenAI · GPT-5.4

  11. GPT-5.3-Codex

    out

    • context 400k tokensContext window of GPT-5.3-Codex. GPT-5.2-Codex is the previous model of the same tier, but the gap is beyond what a percentage explains.
    • price $1.75 / MTokPrompt price of GPT-5.3-Codex. GPT-5.2-Codex is the previous model of the same tier, but the gap is beyond what a percentage explains.

    GPT-5.3-Codex is OpenAI’s most advanced agentic coding model, combining the frontier software engineering performance of GPT-5.2-Codex with the broader reasoning and professional knowledge capabilities of GPT-5.2. It achieves state-of-the-art results...

    OpenAI · GPT-5.3-Codex

  12. GPT-5.2-Codex

    out

    • context +213 %Context window compared with GPT-5.2 Chat, same tier in this family. Went from 128k tokens to 400k tokens.
    • price $1.75 / MTokPrompt price of GPT-5.2-Codex. GPT-5.2 Chat is the previous model of the same tier, but the gap is beyond what a percentage explains.

    GPT-5.2-Codex is an upgraded version of GPT-5.1-Codex optimized for software engineering and coding workflows. It is designed for both interactive development sessions and long, independent execution of complex engineering tasks....

    OpenAI · GPT-5.2-Codex

  13. 2 models on this date

    GPT-5.2 Chat

    out

    • context -68 %Context window compared with GPT-5.1, same tier in this family. Went from 400k tokens to 128k tokens.
    • price +40 %Prompt price compared with GPT-5.1, same tier in this family. Went from $1.25 / MTok to $1.75 / MTok. Lower is better.

    GPT-5.2 Chat (AKA Instant) is the fast, lightweight member of the 5.2 family, optimized for low-latency chat while retaining strong general intelligence. It uses adaptive reasoning to selectively “think” on...

    OpenAI · GPT-5.2

    GPT-5.2

    out

    • context 400k tokensContext window of GPT-5.2. GPT-5.1 is the previous model of the same tier, but the gap is beyond what a percentage explains.
    • price +40 %Prompt price compared with GPT-5.1, same tier in this family. Went from $1.25 / MTok to $1.75 / MTok. Lower is better.

    GPT-5.2 is the latest frontier-grade model in the GPT-5 series, offering stronger agentic and long context perfomance compared to GPT-5.1. It uses adaptive reasoning to allocate computation dynamically, responding quickly...

    OpenAI · GPT-5.2

  14. GPT-5.1-Codex-Max

    out

    • context 400k tokensContext window of GPT-5.1-Codex-Max. First tracked model of this family in the catalogue.
    • price $1.25 / MTokPrompt price of GPT-5.1-Codex-Max. First tracked model of this family in the catalogue.

    GPT-5.1-Codex-Max is OpenAI’s latest agentic coding model, designed for long-running, high-context software development tasks. It is based on an updated version of the 5.1 reasoning stack and trained on agentic...

    OpenAI · GPT-5.1-Codex-Max

  15. GPT-5.1-Codex

    out

    • context +205 %Context window compared with gpt-oss-safeguard-20b, same tier in this family. Went from 131k tokens to 400k tokens.
    • price $1.25 / MTokPrompt price of GPT-5.1-Codex. gpt-oss-safeguard-20b is the previous model of the same tier, but the gap is beyond what a percentage explains.

    GPT-5.1-Codex is a specialized version of GPT-5.1 optimized for software engineering and coding workflows. It is designed for both interactive development sessions and long, independent execution of complex engineering tasks....

    OpenAI · GPT-5.1 for developers

  16. GPT-5.1

    out

    • context +205 %Context window compared with gpt-oss-safeguard-20b, same tier in this family. Went from 131k tokens to 400k tokens.
    • price $1.25 / MTokPrompt price of GPT-5.1. gpt-oss-safeguard-20b is the previous model of the same tier, but the gap is beyond what a percentage explains.

    GPT-5.1 is the latest frontier-grade model in the GPT-5 series, offering stronger general-purpose reasoning, improved instruction adherence, and a more natural conversational style compared to GPT-5. It uses adaptive reasoning...

    OpenAI · GPT-5.1

  17. gpt-oss-safeguard-20b

    out

    • context -67 %Context window compared with GPT-5, same tier in this family. Went from 400k tokens to 131k tokens.
    • price -94 %Prompt price compared with GPT-5, same tier in this family. Went from $1.25 / MTok to $0.075 / MTok. Lower is better.

    gpt-oss-safeguard-20b is a safety reasoning model from OpenAI built upon gpt-oss-20b. This open-weight, 21B-parameter Mixture-of-Experts (MoE) model offers lower latency for safety tasks like content classification, LLM filtering, and trust...

    OpenAI · gpt-oss-safeguard

    open weights · Hugging Face · 18,847 downloads

  18. GPT-5 Pro

    catalogued

    Release date unconfirmed

    • context 400k tokensContext window of GPT-5 Pro. First tracked model of this family in the catalogue.
    • price $15.00 / MTokPrompt price of GPT-5 Pro. First tracked model of this family in the catalogue.

    GPT-5 Pro is OpenAI’s most advanced model, offering major improvements in reasoning, code quality, and user experience. It is optimized for complex tasks that require step-by-step reasoning, instruction following, and...

    OpenRouter catalogue

  19. GPT-5

    out

    • context +205 %Context window compared with gpt-oss-120b, same tier in this family. Went from 131k tokens to 400k tokens.
    • price $1.25 / MTokPrompt price of GPT-5. gpt-oss-120b is the previous model of the same tier, but the gap is beyond what a percentage explains.

    GPT-5 is OpenAI’s most advanced model, offering major improvements in reasoning, code quality, and user experience. It is optimized for complex tasks that require step-by-step reasoning, instruction following, and accuracy...

    OpenAI · GPT-5 for developers

  20. 2 models on this date

    gpt-oss-120b

    out

    • context -87 %Context window compared with GPT-4.1, same tier in this family. Went from 1.05M tokens to 131k tokens.
    • price -92 %Prompt price compared with GPT-4.1, same tier in this family. Went from $2.00 / MTok to $0.150 / MTok. Lower is better.

    gpt-oss-120b is an open-weight, 117B-parameter Mixture-of-Experts (MoE) language model from OpenAI designed for high-reasoning, agentic, and general-purpose production use cases. It activates 5.1B parameters per forward pass and is optimized...

    OpenAI · gpt-oss

    open weights · Hugging Face · 2,475,702 downloads

    gpt-oss-20b

    out

    • context -87 %Context window compared with GPT-4.1, same tier in this family. Went from 1.05M tokens to 131k tokens.
    • price -98 %Prompt price compared with GPT-4.1, same tier in this family. Went from $2.00 / MTok to $0.030 / MTok. Lower is better.

    gpt-oss-20b is an open-weight 21B parameter model released by OpenAI under the Apache 2.0 license. It uses a Mixture-of-Experts (MoE) architecture with 3.6B active parameters per forward pass, optimized for...

    OpenAI · gpt-oss

    open weights · Hugging Face · 6,046,280 downloads

  21. GPT-4.1

    out

    • context 1.05M tokensContext window of GPT-4.1. GPT-4o (2024-11-20) is the previous model of the same tier, but the gap is beyond what a percentage explains.
    • price -20 %Prompt price compared with GPT-4o (2024-11-20), same tier in this family. Went from $2.50 / MTok to $2.00 / MTok. Lower is better.

    GPT-4.1 is a flagship large language model optimized for advanced instruction following, real-world software engineering, and long-context reasoning. It supports a 1 million token context window and outperforms GPT-4o and...

    OpenAI · GPT-4.1 in the API

  22. GPT-4o (2024-11-20)

    catalogued

    Release date unconfirmed

    • context 128k tokensContext window of GPT-4o (2024-11-20). GPT-4o (2024-08-06) is the previous model of the same tier, but the gap is beyond what a percentage explains.
    • price $2.50 / MTokPrompt price of GPT-4o (2024-11-20). GPT-4o (2024-08-06) is the previous model of the same tier, but the gap is beyond what a percentage explains.

    The 2024-11-20 version of GPT-4o offers a leveled-up creative writing ability with more natural, engaging, and tailored writing to improve relevance & readability. It’s also better at working with uploaded...

    OpenRouter catalogue

  23. GPT-4o (2024-08-06)

    out

    • context 128k tokensContext window of GPT-4o (2024-08-06). GPT-4o is the previous model of the same tier, but the gap is beyond what a percentage explains.
    • price $2.50 / MTokPrompt price of GPT-4o (2024-08-06). GPT-4o is the previous model of the same tier, but the gap is beyond what a percentage explains.

    The 2024-08-06 version of GPT-4o offers improved performance in structured outputs, with the ability to supply a JSON schema in the respone_format. Read more here.

    OpenAI · Structured Outputs in the API

  24. 2 models on this date

    GPT-4o-mini

    out

    • context 128k tokensContext window of GPT-4o-mini. First tracked model of this family in the catalogue.
    • price $0.150 / MTokPrompt price of GPT-4o-mini. First tracked model of this family in the catalogue.

    GPT-4o mini is OpenAI's newest model after GPT-4 Omni, supporting both text and image inputs with text outputs. As their most advanced small model, it is many multiples more affordable...

    OpenAI · GPT-4o mini

    GPT-4o-mini (2024-07-18)

    out

    • context 128k tokensContext window of GPT-4o-mini (2024-07-18). First tracked model of this family in the catalogue.
    • price $0.150 / MTokPrompt price of GPT-4o-mini (2024-07-18). First tracked model of this family in the catalogue.

    GPT-4o mini is OpenAI's newest model after GPT-4 Omni, supporting both text and image inputs with text outputs. As their most advanced small model, it is many multiples more affordable...

    OpenAI · GPT-4o mini

  25. 2 models on this date

    GPT-4o

    out

    • context 128k tokensContext window of GPT-4o. GPT-4 is the previous model of the same tier, but the gap is beyond what a percentage explains.
    • price -92 %Prompt price compared with GPT-4, same tier in this family. Went from $30.00 / MTok to $2.50 / MTok. Lower is better.

    GPT-4o ("o" for "omni") is OpenAI's latest AI model, supporting both text and image inputs with text outputs. It maintains the intelligence level of GPT-4 Turbo while being twice as...

    OpenAI · Hello GPT-4o

    GPT-4o (2024-05-13)

    out

    • context 128k tokensContext window of GPT-4o (2024-05-13). GPT-4 is the previous model of the same tier, but the gap is beyond what a percentage explains.
    • price -83 %Prompt price compared with GPT-4, same tier in this family. Went from $30.00 / MTok to $5.00 / MTok. Lower is better.

    GPT-4o ("o" for "omni") is OpenAI's latest AI model, supporting both text and image inputs with text outputs. It maintains the intelligence level of GPT-4 Turbo while being twice as...

    OpenAI · Hello GPT-4o

  26. GPT-4 Turbo

    catalogued

    Release date unconfirmed

    • context 128k tokensContext window of GPT-4 Turbo. GPT-3.5 Turbo (older v0613) is the previous model of the same tier, but the gap is beyond what a percentage explains.
    • price $10.00 / MTokPrompt price of GPT-4 Turbo. GPT-3.5 Turbo (older v0613) is the previous model of the same tier, but the gap is beyond what a percentage explains.

    The latest GPT-4 Turbo model with vision capabilities. Vision requests can now use JSON mode and function calling.

    OpenRouter catalogue

  27. GPT-3.5 Turbo Instruct

    catalogued

    Release date unconfirmed

    • context -75 %Context window compared with GPT-3.5 Turbo 16k, same tier in this family. Went from 16k tokens to 4k tokens.
    • price -50 %Prompt price compared with GPT-3.5 Turbo 16k, same tier in this family. Went from $3.00 / MTok to $1.50 / MTok. Lower is better.

    This model is a variant of GPT-3.5 Turbo tuned for instructional prompts and omitting chat-related optimizations. Training data: up to Sep 2021.

    OpenRouter catalogue

  28. 2 models on this date

    GPT-3.5 Turbo (older v0613)

    out

    • context 4k tokensContext window of GPT-3.5 Turbo (older v0613). GPT-3.5 Turbo Instruct is the previous model of the same tier, but the gap is beyond what a percentage explains.
    • price -33 %Prompt price compared with GPT-3.5 Turbo Instruct, same tier in this family. Went from $1.50 / MTok to $1.00 / MTok. Lower is better.

    GPT-3.5 Turbo is OpenAI's fastest model. It can understand and generate natural language or code, and is optimized for chat and traditional completion tasks.

    OpenAI · Function calling and API updates

    GPT-3.5 Turbo 16k

    out

    • context 16k tokensContext window of GPT-3.5 Turbo 16k. GPT-3.5 Turbo is the previous model of the same tier, but the gap is beyond what a percentage explains.
    • price $3.00 / MTokPrompt price of GPT-3.5 Turbo 16k. GPT-3.5 Turbo is the previous model of the same tier, but the gap is beyond what a percentage explains.

    This model offers four times the context length of gpt-3.5-turbo, allowing it to support approximately 20 pages of text in a single request at a higher cost. Training data: up...

    OpenAI · Function calling and API updates

  29. GPT-4

    out

    • context 8k tokensContext window of GPT-4. First tracked model of this family in the catalogue.
    • price $30.00 / MTokPrompt price of GPT-4. First tracked model of this family in the catalogue.

    OpenAI's flagship model, GPT-4 is a large-scale multimodal language model capable of solving difficult problems with greater accuracy than previous models due to its broader general knowledge and advanced reasoning...

    OpenAI · GPT-4

  30. GPT-3.5 Turbo

    out

    • context 16k tokensContext window of GPT-3.5 Turbo. First tracked model of this family in the catalogue.
    • price $0.500 / MTokPrompt price of GPT-3.5 Turbo. First tracked model of this family in the catalogue.

    GPT-3.5 Turbo is OpenAI's fastest model. It can understand and generate natural language or code, and is optimized for chat and traditional completion tasks.

    OpenAI · ChatGPT and Whisper APIs

  31. GPT-3

    catalogued

    Historical reference

    In May 2020, GPT-3 turned few-shot prompting into a product. Public date from the family biography or the public catalogue; a listing date is not always the release day.

    arXiv · Language Models are Few-Shot Learners

  32. GPT-2

    catalogued

    Historical reference

    GPT-2 followed in February 2019, released in stages because the lab judged the text too convincing. The full 1.5B model came in November 2019. Public date from the family biography or the public catalogue; a listing date is not always the release day.

    OpenAI · GPT-2 paper

  33. GPT-1

    catalogued

    Historical reference

    OpenAI opened the line with GPT-1, a transformer trained to show that generative pre-training worked. First paper on that idea. Public date from the family biography or the public catalogue; a listing date is not always the release day.

    OpenAI · GPT-1 paper

the lab's story

the story

OpenAI started small in 2018 with GPT-1, just to see if a machine could learn to write on its own. A year later GPT-2 already wrote too well, so the lab released it in stages. In 2020, GPT-3 put that in everyone's hands. GPT-4 added images in 2023. GPT-5 came in 2025, then GPT-6 Astra on 3 September 2026, today's flagship. GPT-6 Sol and Luna followed on 22 September, and GPT-6.1 Sol on 29 September, at DevDay.

  1. 11 Jun 2018GPT-1
  2. 14 Feb 2019GPT-2
  3. 28 May 2020GPT-3
  4. 14 Mar 2023GPT-4
  5. 7 Aug 2025GPT-5
  6. 3 Sep 2026GPT-6 Astra
  7. 29 Sep 2026GPT-6.1 Sol

From the lab's public notes.