The AI model release tracker

Next AI Model follows every release from OpenAI, Google, Anthropic, xAI, Meta, Mistral, Alibaba, DeepSeek, Microsoft, Moonshot (Kimi), Z.ai (GLM), Xiaomi (MiMo), ByteDance (Seed, Seedream, Seedance), Black Forest Labs (FLUX), NVIDIA (Nemotron), MiniMax (MiniMax M, Hailuo) and Kuaishou (Kling): what each new model changed, in the maker's own numbers, and which lines are due next by their own release rhythm.

Checked for new releases Oct 19, 2026

Coming next

Every AI model line we track, grouped by where it stands in its own release rhythm, likeliest to release first; within a group, bigger makers first. Each card compares the days since the last release with the line's usual gap. Closed models and open-weight models are listed separately. What the labels mean

Announced1 lineThe maker has announced the next model; it is not open to everyone yet.
xAI

Grok

xAI · flagship line

Latest: Grok 4.7 · Sep 21, 2026

AnnouncedGrok 4.8 · announced Sep 14, 2026, 35 days ago

“Grok 4.8, which is a 2.5T model trained with our new C++ software stack, will finish training this week and start RL” — xAI, Sep 14, 2026 · Elon Musk on X

A larger base model than Grok 4.6, trained longer on multi-hour agentic tasks, at the same $2 / $6 price and the same 500k-token context.

What changed in Grok 4.7 →
Days since last release
28
days
Usual gap
59
days
Recent gaps
Grok 4.7 · Sep 2026
40
Grok 4.6 · Aug 2026
35
Grok 4.5 · Jul 2026
82
Grok 4.3 · Apr 2026
59
Longer than usual2 linesPast their usual gap. About 1 in 4 got a new model within a month, most often soon after the gap: those come first.

Seed

ByteDance · flagship line

Latest: Doubao-Seed-2.1-Pro · Jun 23, 2026

Longer than usual118 days since the last release · usual gap 85 days
Days since last release
118
days
Usual gap
85
days
Recent gaps
Doubao-Seed-2.1-Pro · Jun 2026
129
Doubao-Seed-2.0-Pro · Feb 2026
58
Doubao-Seed-1.8 · Dec 2025
190
Doubao-Seed-1.6 · Jun 2025
55

Muse Spark

Meta · flagship line

Latest: Muse Spark 1.3 · Sep 2, 2026

Longer than usual47 days since the last release · usual gap 28 days

Five points up on the all-round intelligence index over Muse Spark 1.2 and higher scores on office and terminal tasks, at the same price and the same 1M-token context.

What changed in Muse Spark 1.3 →
Days since last release
47
days
Usual gap
28
days
Recent gaps
Muse Spark 1.3 · Sep 2026
28
Muse Spark 1.2 · Aug 2026
27
Muse Spark 1.1 · Jul 2026
92
Mid-cycle2 linesPartway through their usual gap. About 1 in 4 got a new model within a month; the ones closest to their gap come first.

Qwen Max

Alibaba · flagship line

Latest: Qwen3.8-Max-0902 · Sep 2, 2026

Mid-cycle47 days since the last release · usual gap 59 days

An upgraded snapshot of Qwen 3.8-Max, further trained for coding and office-style agent work, with the same 2.4 trillion parameters, 1M-token context window and $2 / $6 price.

What changed in Qwen3.8-Max-0902 →
Days since last release
47
days
Usual gap
59
days
Recent gaps
Qwen3.8-Max-0902 · Sep 2026
45
Qwen 3.8-Max · Jul 2026
60
Qwen 3.7-Max · May 2026
32
Qwen3.6-Max-Preview · Apr 2026
166

38 lines closed by the maker or without a release in over a year: see every line we track.

Get the weekly email

Every Monday, only when something shipped: the AI models that came out, what each one changed, and which lines are due next.

We email you a link to confirm first · unsubscribe anytime · read past issues · about the newsletter · RSS feed

Fresh off the press

Just dropped

The latest releases, newest first. Open a model to see what changed, in numbers, against the model it replaces.

Oct
17
2026

An update to DeepSeek-V4.1-Flash that scores higher on every agentic, coding and reasoning test DeepSeek published while giving shorter answers, with the same MIT weights, 1M-token context and API price.

BFL
Oct
15
2026

Brings FLUX 3 Image's editing, up to ten reference images and bounding-box layout to a 5.6B open-weights model that runs on consumer GPUs down to an RTX 3090, with output up to 4K.

Oct
15
2026

Meta's first downloadable Muse Spark: an 816B-parameter mixture-of-experts model with about 42B active parameters, a 1M-token context and text, image, video and audio input, released under the Muse Model License.

Oct
15
2026

An open-weight model with 120B parameters, 12B active at a time, that reads text, images and videos up to 2 minutes long, with a context of up to 1M tokens.

Oct
13
2026

Google's first Gemini 4 model replaces Gemini 3.1 Pro as its top model in the paid Gemini API, with a 1M-token window shared by input, thinking and output.

Oct
12
2026

Meta's first frontier-tier model, a larger step above Muse Spark built for agents, scoring 4 points higher on the Artificial Analysis index at about three to five times the API price.

Questions people ask

What is the next OpenAI model?

OpenAI's most recent flagship release is GPT-6.1 Sol (GPT line, September 29, 2026); its other flagship line's latest model is GPT-6 Astra (GPT Astra line, September 4, 2026). It shipped 20 days ago, and the line's usual gap is about 75 days. We only report official announcements, never rumored release dates.

When is the next Claude model coming out?

Anthropic's most recent flagship release is Claude Opus 5.5 (Claude Opus line, September 22, 2026); its other flagship line's latest model is Claude Fable 5.1 (Claude Fable line, September 1, 2026). The line's usual gap is about 60 days and 27 have passed, so it is mid-cycle. We only report official announcements, never rumored release dates.

When is the next Gemini model coming out?

Google's newest flagship model is Gemini 4 Argon, released on October 13, 2026. It shipped 6 days ago, and the line's usual gap is about 236 days. We only report official announcements, never rumored release dates.

What is the newest AI model?

The most recent releases we track are DeepSeek-V4.1-Flash-1017 by DeepSeek (Oct 17, 2026), Muse Spark Open by Meta (Oct 15, 2026) and Nemotron 3.5 Super VL by NVIDIA (Oct 15, 2026). The archive lists every release, newest first.

How does Next AI Model decide that a model line is due?

Each line's usual gap is the median of its last five gaps between releases, counting launches less than a week apart as one. A line is Due from 85% to 100% of that gap, Longer than usual after that, Mid-cycle in between, and Just released during the first third. In our archive, about 1 in 3 Due lines got a new model within a month, about twice the average. Lines with fewer than two gaps between releases are marked “Too early to call”, because their usual gap is not known yet, and lines with exactly two say “from only 2 gaps”, because such a short history is only a rough guide. It is a reading of the line's own rhythm, not insider information.

Why doesn't Next AI Model give exact release dates?

We tested a simple forecast, the usual gap plus or minus 15 days, against every release in our archive: it would have been right fewer than 1 time in 5. Release rhythms are too irregular for a date to be honest, so we show where each line stands instead.