Skip to content
FrontierDot

AI model releases in 2026

Each row is one node from a vendor timeline already published here. This page does not add a release that those timelines do not list. The date and the wording come from that node. The source link is the page that timeline cites.

Other years: 2024 2025

Date Vendor What the timeline says Source cited there
2026-02-05 Anthropic timeline Claude Opus 4.6 launches for complex agentic and long-horizon work. The release notes say it recommends adaptive thinking and that manual thinking budgets are deprecated. Anthropic: Claude Opus 4.6
2026-02-17 Anthropic timeline Claude Sonnet 4.6 launches as the balanced everyday model, with extended thinking and a 1M-token context window in beta according to the release notes. Anthropic: Introducing Sonnet 4.6
2026-02-19 Google timeline Gemini 3.1 Pro Preview is released, along with a separate custom-tools endpoint. Gemini API changelog
2026-02-24 OpenAI timeline gpt-5.3-codex, a coding-specialized model, is released to the Responses API. OpenAI API changelog
2026-03-03 Google timeline Gemini 3.1 Flash-Lite Preview launches as the first Flash-Lite model in the Gemini 3 series. Gemini API changelog
2026-03-03 OpenAI timeline GPT-5.3 Instant reaches the API as gpt-5.3-chat-latest, the chat model used in ChatGPT. OpenAI API changelog
2026-03-05 OpenAI timeline GPT-5.4 and GPT-5.4 Pro are released. OpenAI: Introducing GPT-5.4
2026-03-10 xAI timeline The xAI release notes record that Grok 4.20 and Grok 4.20 Multi-agent are live in the API. xAI developer release notes
2026-03-17 OpenAI timeline GPT-5.4 mini and GPT-5.4 nano are released in the API. OpenAI API changelog
2026-04-07 Anthropic timeline Anthropic announces Claude Mythos Preview, a model it calls strikingly capable at computer security, and does not release it generally. It goes instead to Project Glasswing, a defensive effort with Amazon Web Services, Apple, Broadcom, Cisco, CrowdStrike, Google, JPMorganChase, the Linux Foundation, Microsoft, NVIDIA and Palo Alto Networks. Anthropic: Project Glasswing
2026-04-08 Meta timeline Meta introduces Muse Spark, the first model in the Muse family from Meta Superintelligence Labs: a natively multimodal reasoning model with tool use, visual chain of thought and multi-agent orchestration. Meta calls it the first step on its scaling ladder. Meta AI: Introducing Muse Spark
2026-04-16 Anthropic timeline Claude Opus 4.7 launches at the same $5 / $25 per million tokens as Opus 4.6, with an updated tokenizer. Anthropic: Introducing Claude Opus 4.7
2026-04-24 OpenAI timeline GPT-5.5 and GPT-5.5 Pro are released in the API. OpenAI API changelog
2026-05-07 Google timeline Gemini 3.1 Flash-Lite becomes generally available. Gemini API changelog
2026-05-19 Google timeline Gemini 3.5 Flash becomes generally available and becomes the model behind gemini-flash-latest. Gemini API changelog
2026-05-28 Anthropic timeline Claude Opus 4.8 launches with a 1M-token context window by default and 128k max output tokens. Anthropic: Introducing Claude Opus 4.8
2026-06-09 Anthropic timeline Claude Fable 5 and Claude Mythos 5 launch. Anthropic says they are the same underlying model: Fable 5 is generally available with safeguards that hand some cyber, biology and distillation requests to Claude Opus 4.8, while Mythos 5 lifts some of those safeguards and goes only to Project Glasswing partners. Both cost $10 / $50 per million tokens. Anthropic: Claude Fable 5 and Claude Mythos 5
2026-06-12 Anthropic timeline Anthropic suspends access to Fable 5 and Mythos 5 for all customers. Its statement says a US government export control directive, received at 5:21pm ET, barred any foreign national from accessing the two models, and that complying meant disabling them for everyone. Other Claude models were not affected. Anthropic: Statement on the directive to suspend Fable 5 access
2026-06-30 Anthropic timeline Claude Sonnet 5 launches at introductory pricing of $2 / $10 per million tokens, made the standard price on August 10. The same day Anthropic publishes “Redeploying Claude Fable 5”. Anthropic: Introducing Claude Sonnet 5
2026-06-30 Google timeline Gemini Omni Flash enters public preview as a multimodal model for fast video generation and conversational video editing, producing 3–10 second clips at 720p. Gemini API changelog
2026-07-01 Anthropic timeline The release notes record that access to Claude Fable 5 and Claude Mythos 5 has been restored. Claude Platform release notes
2026-07-08 xAI timeline Grok 4.5 is announced, priced at $2 per million input tokens and $6 per million output tokens. xAI: Grok 4.5
2026-07-09 Meta timeline Muse Spark 1.1 is released as a significant upgrade, with gains in tool and computer use, coding and multimodal understanding. Developers can use it through the Meta Model API, now in public preview. Meta AI: Introducing Muse Spark 1.1
2026-07-09 OpenAI timeline The GPT-5.6 family is released in three sizes named Sol, Terra and Luna, the naming OpenAI carries into GPT-6. OpenAI API changelog
2026-07-21 Google timeline Gemini 3.6 Flash and Gemini 3.5 Flash-Lite become generally available; the changelog says 3.6 Flash is cheaper than 3.5 Flash with better token efficiency. Gemini API changelog
2026-07-22 xAI timeline Grok 4.5 starts powering Grok on grok.com, on X and in the iOS and Android apps. xAI: Grok 4.5 everywhere
2026-07-24 Anthropic timeline Claude Opus 5 launches at $5 / $25 per million tokens with a 1M-token context window and 128k max output tokens. Anthropic: Introducing Claude Opus 5
2026-07-28 xAI timeline Elon Musk posts: “Grok 4.6 releases around August 7. This will be the 1.5T model with significantly improved SFT & RL.” In the same post: “Grok 4.7 will be the 2.1T model released a few weeks later.” Elon Musk on X
2026-07-30 xAI timeline Musk posts: “And Grok 4.6 comes out in a week”. Elon Musk on X
2026-08 Meta timeline Muse Glimmer, a small dense Muse model meant for running locally, is released this month according to Hugging Face. We found no official page with a day, so this node carries only the month. Hugging Face: Welcome Muse Spark Open
2026-08-05 Meta timeline Meta releases Muse Spark 1.2 and an early beta of Muse Code, a coding agent for long-horizon, multi-agent work whose subagents, tool calls, steers and cancels are recorded in a replayable event log. Meta for Developers: Build with Muse Code
2026-08-12 xAI timeline Grok 4.6 is released, five days after the “around August 7” date Musk had given. xAI: Grok 4.6
2026-08-13 Google timeline Gemini 3.7 Flash becomes generally available, described as Google’s most intelligent workhorse model yet for coding and agents, at an introductory price. Gemini API changelog
2026-08-15 xAI timeline Musk replies: “Worth trying Grok 4.6. Fable 5 is smarter overall, but Grok is much faster and lower cost. Grok 4.7 has a good chance of exceeding all current models in intelligence.” Elon Musk on X
2026-08-19 xAI timeline Grok 4.6 becomes generally available on Amazon Bedrock at $2 input, $0.50 cached input and $6 output per million tokens, with a 500k context window. xAI: Grok 4.6 on Amazon Bedrock
2026-09-01 Anthropic timeline Claude Fable 5.1 and Claude Mythos 5.1 launch, again one model with two safeguard levels. Prices stay at $10 / $50, but cache reads drop to $0.25 per million tokens, which Anthropic estimates cuts typical costs by about 25%. Anthropic: Claude Fable 5.1 and Mythos 5.1
2026-09-01 OpenAI timeline OpenAI publishes “The path to Astra”, saying its next model is the first it treats as reaching the Critical cybersecurity capability level under its Preparedness Framework, and describing a staged rollout. OpenAI: The path to Astra
2026-09-02 Google timeline Gemini 3.8 Flash becomes generally available as Google’s most intelligent Flash model, aimed at long-horizon software engineering and autonomous agents. Gemini API changelog
2026-09-02 Meta timeline Muse Spark 1.3 becomes available in Muse Code and the Meta Model API. Meta says it uses about 20% fewer tool calls and about 25% fewer tokens. Meta Research: Introducing Muse Spark 1.3
2026-09-02 xAI timeline Musk posts: “Grok 4.7 comes out in 10 days”. Elon Musk on X
2026-09-03 OpenAI timeline GPT-6 Astra is released at $10 / $50 per million input and output tokens. It goes first to a limited set of organizations, then to ChatGPT Plus, Pro, Business and Enterprise, the API, Microsoft Azure and Amazon Bedrock. OpenAI: GPT-6 Astra
2026-09-11 xAI timeline Musk posts: “Grok 4.7 needs a few more days to cook.” He adds: “We might have penalized response length too much (or something) in RL, as it still gives up on hard tasks (that it can do!) too early and isn’t yet sufficiently rigorous in checking its work.” Elon Musk on X
2026-09-14 xAI timeline Musk posts: “Grok 4.7 should be roughly on par with Opus 5.0, not 5.1. Better in some ways, worse in others. We need to fix multimodal performance. Grok 4.8 will be a noticeable improvement. Grok 4.9 is probably Astra/Fable class. Grok 5 maybe better than anything. We shall see.” Elon Musk on X
2026-09-21 xAI timeline Grok 4.7 is released. SpaceXSI’s post reads: “Grok 4.7 is here. It's a notable improvement over Grok 4.6 at the same price and speed.” SpaceXSI on X: Grok 4.7 is here
2026-09-22 Anthropic timeline Claude Opus 5.5 launches at $4 / $20 per million tokens, below the $5 / $25 of Opus 5. Anthropic: Introducing Claude Opus 5.5
2026-09-22 OpenAI timeline GPT-6 Sol and GPT-6 Luna are released. OpenAI prices Sol at $2 / $10 and Luna at $0.10 / $0.50 per million tokens, 50% below the GPT-5.6 equivalents. OpenAI: Introducing GPT-6 Sol and GPT-6 Luna
2026-09-28 Anthropic timeline Claude Sonnet 5.5 launches on the Claude API, Amazon Bedrock, Claude Platform on AWS, Google Cloud and Microsoft Foundry. Anthropic: Introducing Claude Sonnet 5.5
2026-09-29 OpenAI timeline GPT-6.1 Sol is released at $2 input, $0.10 cached input and $10 output per million tokens. The same day the changelog adds an Ultrafast mode for GPT-6 Astra in the API. OpenAI: Introducing GPT-6.1 Sol
2026-09-30 Google timeline Google announces Gemini 4 Argon and first makes it available to trusted cyber defenders through its Fairwind Program. Google: Gemini 4 Argon for trusted cyber defenders
2026-10-06 Anthropic timeline Anthropic expands its Cyber Verification Program into three access tiers, each including Claude Opus 5.5, Claude Sonnet 5.5 and Claude Mythos 5.1. Existing Project Glasswing members move to the Specialized Access tier. Anthropic: Expanding the Cyber Verification Program
2026-10-07 Anthropic timeline Claude Haiku 5.5 launches for high-volume and latency-sensitive work, with a 1M-token context window and 128k max output tokens. Anthropic: Introducing Claude Haiku 5.5
2026-10-07 OpenAI timeline The ChatGPT release notes announce GPT-6 with Intelligent UI in ChatGPT. OpenAI Help Center: ChatGPT release notes
2026-10-07 xAI timeline Musk posts: “However, most @Bot requests are pretty simple and will be handled by a lightning-fast version of Grok 4.8 when that comes out.” Elon Musk on X
2026-10-08 Google timeline Gemini 3.7 Flash is deprecated; API requests to gemini-3.7-flash are automatically routed to gemini-3.8-flash. Gemini API changelog
2026-10-08 OpenAI timeline Ultrafast mode is added for GPT-6.1 Sol in the Responses API, available to all API users subject to rate limits. OpenAI API changelog
2026-10-12 Meta timeline Meta launches Muse Wonder on Muse Code and the Meta Model API at $4 input, $0.20 cached input and $20 output per million tokens, with a 1M-token context. Muse Wonder Full, Meta’s most advanced model, is available only by subscription in Muse Code on the Power Usage plan and a new Maximum Usage plan with 80x the Everyday Usage limit. Meta for Developers: Put your ideas to work with Muse Wonder
2026-10-13 Google timeline Gemini 4 Argon becomes generally available, described as Google’s most capable model for complex coding, agentic, math and science tasks. Google: Introducing Gemini 4 Argon
2026-10-14 OpenAI timeline GPT-5.5 is retired from ChatGPT, according to OpenAI’s model release notes. OpenAI Help Center: Model release notes
2026-10-15 Meta timeline Meta releases the weights of Muse Spark Open, an 816B-total, 42B-active model with a 1M-token context under the Muse Model License. Meta Research: Introducing Muse Spark Open
2026-10-20 OpenAI timeline The ChatGPT release notes announce an updated GPT-6 Pro that can run parallel agents. OpenAI Help Center: ChatGPT release notes
2026-10-22 Meta timeline Paradigm, with Meta and Andromeda, launches the Muse Spark Open Kernel Challenge: a $100,000 prize pool to speed Muse Spark Open on H200 GPUs, with kernel and serving tracks ending November 19. Paradigm: Muse Spark Open Kernel Challenge

Sources

Last checked 2026-10-22. Rows are assembled from the vendor timeline posts on this site. Those posts cite each lab’s own announcement. This index does not add dates that are not in those posts.