The waiting board
Tracking 21 models.
OutYet.ai tracks upcoming AI model releases from OpenAI, Anthropic, Google DeepMind, Meta, and the other major labs. Follow whatever you're waiting on and get an alert when public access is verified.
One click creates your free account (or signs you into an existing one) and turns on alerts for every major release. There's no password to set, and you can unsubscribe any time.
21 models trackedTwo verified checks before any alertNo spam, leave anytime
Is Mistral Code Agent Latest out yet?
Yes
Released Jul 31, 2026
Mistral Code Agent Latest is a language model from Mistral AI, now publicly released. It is positioned as the latest version of the company's code agent offering.
Is Claude Opus 5 out yet?
Yes
Released Jul 24, 2026
Claude Opus 5 is Anthropic's latest language model, now publicly released. It is the newest entry in the company's Opus line of Claude models.
Is GPT-5.6 Luna out yet?
Yes
Released Jul 9, 2026
OpenAI's GPT-5.6 Luna model, tracked with official availability evidence and vendor-reported HealthBench results.
Is GPT-5.6 Sol out yet?
Yes
Released Jul 9, 2026
OpenAI's next-generation model with stronger capabilities in coding, science, and cybersecurity, paired with an advanced safety stack.
Is GPT-5.6 Terra out yet?
Yes
Released Jul 9, 2026
OpenAI's GPT-5.6 Terra model, tracked with official availability evidence and vendor-reported HealthBench results.
Is GPT-Live out yet?
Yes
Released Jul 9, 2026
A new generation of voice models for natural human-AI interaction now powering ChatGPT Voice.
Is Grok 4.5 out yet?
Yes
Released Jul 8, 2026
Grok 4.5 is an AI model from xAI, now publicly released. Prior to release, it entered private beta testing at Tesla and SpaceX; detailed information about its capabilities has not been provided in official announcements.
Is Leanstral 1.5 out yet?
Yes
Released Jul 2, 2026
Mistral's automated theorem proving model for Lean 4 proof assistance.
Is Claude Fable 5 out yet?
Yes
Released Jul 1, 2026
Claude Fable 5 is now publicly available after its suspension in June 2026 was lifted. The model returns to public access following a reassessment of national security considerations.
Is Claude Mythos 5 out yet?
Yes
Released Jul 1, 2026
Watching for Anthropic to restore the bleeding edge Mythos model.
Is Claude Sonnet 5 out yet?
Yes
Released Jun 30, 2026
Claude Sonnet 5 is Anthropic's most agentic Sonnet model, capable of planning, using tools such as browsers and terminals, and running autonomously at a level approaching Opus 4.8 at lower cost. It improves over Sonnet 4.6 in reasoning, tool use, coding, and knowledge work, shows fewer undesirable safety behaviors, and is available across all Claude plans as the default for Free and Pro, with introductory pricing of $2 per million input tokens and $10 per million output tokens through August 31, 2026.
Is Nano Banana 2 Lite out yet?
Yes
Released Jun 29, 2026
Google's lightweight generative-media (image) model in the Nano Banana line, made available to build with on 2026-06-30.
Is GLM-5.2 out yet?
Yes
Released Jun 13, 2026
Open-weight (MIT) 744B-total / 40B-active flagship coding model with a 1M-token context.
Is Kimi K2.7 Code out yet?
Yes
Released Jun 12, 2026
Coding-specialized release that cut reasoning-token use by ~30%.
Is North Mini Code out yet?
Yes
Released Jun 11, 2026
Cohere's first developer-focused model — an open-weight 30B MoE agentic coder that runs on a single H100.
Is MiniMax M3 out yet?
Yes
Released Jun 1, 2026
MiniMax's June 2026 flagship; weights followed the announcement in mid-June.
Is Claude Opus 4.8 out yet?
Yes
Released May 28, 2026
The current flagship Claude Opus, leading on coding and agentic tasks.
Is Gemini 3.5 Flash out yet?
Yes
Released May 19, 2026
Google's faster Gemini 3.5 sibling to the tracked 3.5 Pro; surfaced via an official post introducing computer use in Gemini 3.5 Flash.
Is Gemini Omni Flash out yet?
Yes
Released May 19, 2026
Google DeepMind's new low-latency multimodal Gemini variant, announced for builders alongside Nano Banana 2 Lite.
Is Qwen 3.7 Max out yet?
Yes
Released May 19, 2026
Alibaba's current flagship, with benchmark wins over frontier closed models.
Is ERNIE 5.1 out yet?
Yes
Released May 8, 2026
Baidu's current flagship — the top Chinese model on LM Arena at launch.
Is GPT-5.4 out yet?
Yes
Released May 5, 2026
A model used in medicinal chemistry research to improve drug-making reactions.
Is GPT-5.5 Instant out yet?
Yes
Released May 5, 2026
A model variant designed to improve health and wellness responses in ChatGPT with stronger reasoning.
Is Granite 4.1 out yet?
Yes
Released Apr 29, 2026
IBM's current family of ten Apache-licensed models across 3B–30B.
Is DeepSeek V4 Flash out yet?
Yes
Released Apr 24, 2026
The fast, low-cost tier of the V4 generation.
Is DeepSeek V4 Pro out yet?
Yes
Released Apr 24, 2026
DeepSeek's MIT-licensed 2026 flagship, leading the open-weight boards.
Is GPT-5.5 out yet?
Yes
Released Apr 23, 2026
OpenAI's spring 2026 flagship, near the top of the LM Arena board at launch.
Is Kimi K2.6 out yet?
Yes
Released Apr 20, 2026
Moonshot's spring 2026 open flagship, near closed-frontier coding quality.
Is Grok 4.3 out yet?
Yes
Released Apr 17, 2026
xAI's current flagship — native video input, document generation, and stronger tool-calling.
Is Muse Spark out yet?
Yes
Released Apr 8, 2026
The first model from Meta Superintelligence Labs — a closed-weight, natively multimodal reasoning model marking Meta's pivot away from open Llama releases.
Is Gemini 3.1 Pro out yet?
Yes
Released Feb 19, 2026
Google's current Pro-tier flagship.
Is Claude Sonnet 4.6 out yet?
Yes
Released Feb 17, 2026
The current balanced Claude Sonnet for high-volume coding and chat.
Is Qwen3-Coder-Next out yet?
Yes
Released Feb 2, 2026
The 2026 refresh of Qwen's open agentic-coding line.
Is Nemotron 3 out yet?
Yes
Released Dec 15, 2025
NVIDIA's current open-weight generation — Nano, Super, and Ultra tiers rolled out December 2025 through June 2026.
Is Amazon Nova 2 Lite out yet?
Yes
Released Dec 2, 2025
The first Nova 2 reasoning model on Bedrock; Pro and Premier tiers are still rolling out.
Is Mistral Large 3 out yet?
Yes
Released Dec 2, 2025
Open-weight 675B sparse MoE flagship (41B active), part of the Mistral 3 launch.
Is OLMo 3 out yet?
Yes
Released Nov 20, 2025
Ai2's current fully open flagship with an inspectable reasoning flow.
Is Claude Haiku 4.5 out yet?
Yes
Released Oct 15, 2025
The fast, low-cost Haiku tier of the Claude 4.5 generation.
Is Seed-OSS 36B out yet?
Yes
Released Aug 20, 2025
ByteDance Seed's Apache-licensed 36B with a native 512K context.
Is Hunyuan-A13B out yet?
Yes
Released Jun 27, 2025
Efficient 80B-total/13B-active open-weight MoE.
Is Magistral out yet?
Yes
Released Jun 10, 2025
Mistral's first dedicated reasoning model line.
Is Llama 4 out yet?
Yes
Released Apr 5, 2025
Meta's mixture-of-experts generation (Scout and Maverick); the previewed Behemoth never shipped.
Is Command A out yet?
Yes
Released Mar 13, 2025
Cohere's efficiency-focused 111B flagship.
Is Gemma 3 out yet?
Yes
Released Mar 12, 2025
Google's current multimodal open-weight family.
Is Claude Mythos 5.1 out yet?
No
Today
3 guesses
Rumored more-capable successor to Claude Mythos 5. Per an internal-portal note (June 22) and AI observer Andrew Curran, Anthropic reportedly finished training it ~9 days after the export-control suspension; may ship as Mythos 5.1 or Mythos 6, or stay internal to accelerate self-improvement. Unconfirmed by Anthropic.
Is Claude Fable 5.1 out yet?
No
Tomorrow
4 guesses
Anticipated successor to Claude Fable 5 — the publicly available counterpart to the rumored Claude Mythos 5.1 / Mythos 6. Unconfirmed by Anthropic.
Is Gemini 3.5 Pro out yet?
No
Today
7 guesses
Google's announced next Pro-tier Gemini model.
Is Gemini 3.5 Flash Cyber out yet?
No
No estimate yet
Google describes Gemini 3.5 Flash Cyber as a lightweight model fine-tuned to find, validate, and patch vulnerabilities. Its initial access is limited to governments and trusted partners through CodeMender.
Is Gemini 3.5 Flash-Lite out yet?
No
No estimate yet
Google documents Gemini 3.5 Flash-Lite as a stable, high-throughput multimodal API model with support for code execution, function calling, search grounding, structured outputs, and thinking.
Is Gemini 3.6 Flash out yet?
No
No estimate yet
Google introduced Gemini 3.6 Flash in a July 21 announcement. The watch records the named model for the existing Google detector; public API availability remains for detector confirmation.
Is Gemini 4 out yet?
No
No estimate yet
Watching Google's official surfaces for the next major Gemini generation beyond the 3.x line. Rumored only — nothing announced yet.
Is Gemini Robotics ER 2 out yet?
No
No estimate yet
Google describes Gemini Robotics ER 2 as an embodied-reasoning model for video understanding, task orchestration, and multi-robot collaboration, with developer access through the Gemini API and AI Studio.
Is Llama 5 out yet?
No
No estimate yet
Watching Meta's official surfaces for a next open-weight Llama generation. Meta's frontier work shifted to the closed Muse models in April 2026, so an open Llama 5 remains unconfirmed.
Is Grok 5 out yet?
No
Today
1 guess
xAI's next flagship; rumored ~6T-parameter MoE with multimodal support and a large context window. Targeted for Q2 2026 but slipping.
Is Cursor foundation model out yet?
No
Today
1 guess
Anysphere's first fully self-trained frontier model, reportedly shipping within weeks.
The model release waiting room
Tracking 21 announced models across 21 providers, plus surprise releases discovered on official model registries.
One click creates your free account (or signs you into an existing one) and turns on alerts for every major release. There's no password to set, and you can unsubscribe any time.
How it works
We check provider APIs and official weight repositories around the clock, and verify access twice before a model's page flips to “Yes”. News and rumors feed the watchlist, but they never mark a release on their own.
The date under each waiting model is the average of visitor guesses, one editable guess per person. And while everything stays “No”, the Discord is where everyone argues about leaks and benchmarks.
Sign in to get an email or a desktop notification the moment a model you follow ships.
About
OutYet.ai checks provider APIs and official weight repositories directly. News tells us what to watch, and nothing gets marked released without verifiable public access.
Frequently asked
We require two successful checks against an official public surface before changing status or sending alerts.
No. Official news helps maintain the watchlist, but a machine-verifiable API, consumer launch, or complete official weights are required.
Follow one named model, every future model from a selected provider, or all releases across the waiting board.
Missing a model, or wishing OutYet did something it doesn't? Tell us what to build.
