# Grok V9 'done training' — what Musk's claim is actually worth

> On 25 May 2026 Musk said xAI's ~1.5T-parameter Grok V9-Medium had finished training.

*An announcement, not a release — and from a company whose timelines keep slipping.*

By WireRead Editorial · WireRead
Canonical: https://wireread.com/news/grok-v9-training-claim-what-its-worth

There is a difference between *a model exists* and *you can use it*, and xAI announcements have a habit of living in the gap between the two. On 25 May, Elon Musk said a new foundation model — **V9-Medium**, reportedly around **1.5 trillion parameters**, roughly triple the current production model — had finished training, with internal evaluations described as positive. A mid-June release was floated in the coverage that followed. By mid-June: no firm date, no model card, no independent benchmarks. The gap is still the gap.

## What 'finished training' actually means

In the public discourse, 'finished training' tends to be conflated with 'released' or 'ready'. It isn't. Training is the first of several phases: the model learns from data, which is compute-intensive and time-bounded — so 'training complete' is a real milestone. What follows is typically weeks or months of evaluation, safety testing, fine-tuning, and the engineering work of turning an artefact into a product. For a model at the scale Musk described — 1.5 trillion parameters — that pipeline is not trivial. A 'finished training' statement says the learning phase is over. It says nothing about whether the resulting model is coherent, safe, aligned, or better than what's already deployed.

> **Key:** **The throughline:** 'finished training' is a statement about an internal artefact, not a benchmarked, released product. The honest status here is: a large model reportedly completed a training run, claims are unverified by any third party, and the shipping flagship is still **Grok 4.3** (current as of June 2026, per xAI's own developer docs).

## How this sits in the xAI product picture

The V9-Medium announcement exists alongside at least one other unshipped major model: *Grok 5*, which has been hyped for considerably longer and which, per Fello AI's tracker as of early June, remained in training with prediction markets openly skeptical of a June ship. That creates an unusual positioning: a company promoting two separate unreleased models simultaneously — neither of which has an independent benchmark to point to — while the current flagship, Grok 4.3, is the model developers can actually access today.

The parameter count of V9-Medium deserves a note. **1.5 trillion parameters** would place it in rarefied company — roughly the scale of the largest reported frontier models — but parameter counts are not reliable proxies for real-world performance. A well-trained, smaller model often outperforms a larger one on the tasks that matter. The claim is notable as a signal of intent and investment; it is not a quality claim without benchmarks to support it.

| Dimension | What Musk claimed | What's verifiable |
| --- | --- | --- |
| **Model name** | V9-Medium | Reported name, not official model card |
| **Parameter count** | ~1.5 trillion | Unverified; no model card published |
| **Training status** | Complete (25 May) | Claimed; no third-party confirmation |
| **Release timeline** | Mid-June 2026 (floated) | No firm date as of mid-June |
| **Evaluation results** | 'Positive' (internal) | Unverified; no benchmark numbers |
| **Current shipped model** | — | Grok 4.3 (xAI developer docs, June 2026) |

> Reports described V9-Medium as targeting a coding leadership position, with a parameter count roughly triple xAI's current production model — and a mid-June release window that slipped without a public update.
> — [Tech Times](https://www.techtimes.com/articles/317328/20260528/grok-ai-new-model-triples-parameter-count-targets-coding-lead-release-expected-mid-june.htm), 2026-05-28

## The pattern, and what to watch

None of this makes V9-Medium fiction. xAI is a real lab that ships real models: Grok 4.3 exists, Grok Imagine 1.5 Preview shipped in June, and the company moves fast. The pattern being observed is not capability fraud — it is a consistent habit of announcing timelines that compress in the press release and expand in reality. Prediction markets, which aggregate the expectations of informed participants who have financial skin in the game, were giving a June ship for Grok 5 notably short odds as of early June. That skepticism is a useful calibration, not a verdict.

> As of early June 2026 Grok 5 remained in training with no confirmed release date, and prediction markets were skeptical of a June ship. The current production flagship was Grok 4.3.
> — [Fello AI](https://felloai.com/all-we-know-so-far-about-grok-5/), 2026-06-05

The actionable read for anyone making model selection or product decisions: V9-Medium is an announcement about an internal training run, not a shipped option. The model to evaluate today is Grok 4.3. When a model card appears — with real benchmark numbers, an API endpoint, and third-party independent testing — that is the moment to form an opinion on V9's quality and fit. Until then, the announcement is worth noting in a timeline and worth not building a roadmap on.

> **Note:** **What to watch.** The signal that V9-Medium has crossed from 'announced' to 'real' is simple: a model card with published benchmark numbers and an API you can call. No model card means no actionable information.

## Key takeaways

- Musk said a new ~1.5T-parameter model, V9-Medium, had finished training on 25 May 2026, with internal evals described as positive.
- A release was reportedly expected mid-June 2026 — but no firm date or model card was published.
- These are unverified claims; xAI's release timelines have slipped repeatedly.
- The separately hyped 'Grok 5' was still in training; the shipping flagship at the time was Grok 4.3.
- A 1.5T-parameter model would be roughly three times the size of the current production model — if the parameter count is accurate.

## FAQ

### Can I use Grok V9-Medium?
Not as of mid-June 2026. Musk said on 25 May that it had finished training, but there was no confirmed release date, model card or independent benchmarks. The model developers can access today is Grok 4.3 (xAI developer docs, June 2026).

### Is this the same as 'Grok 5'?
No. Grok 5 is a separately hyped, larger model that reportedly remained in training as of early June 2026. V9-Medium is a distinct ~1.5T-parameter model Musk said had finished training on 25 May. Both claims are unverified, and xAI's shipping flagship at the time was Grok 4.3.

### Why should I be skeptical of the mid-June release date?
xAI's release timelines have slipped before. By mid-June 2026 — the floated window — no firm date, model card or independent benchmarks had appeared. Prediction markets were also skeptical of the separate 'Grok 5' June ship date, per Fello AI's tracker (June 2026).

### What does '1.5 trillion parameters' mean in practice?
A very large model by current standards — roughly three times xAI's production model and comparable to the largest reported frontier models. However, raw parameter count is not a reliable quality signal: benchmark performance, training quality and fine-tuning often matter more. The figure is unverified and no independent testing exists as of mid-June.

### What is xAI's current best model I can actually use?
Grok 4.3, which is documented on xAI's developer platform (docs.x.ai, June 2026) as the current production flagship. It is the model available via the API until V9-Medium or Grok 5 ships with a model card.

## Sources

- [Grok AI New Model Triples Parameter Count, Targets Coding Lead: Release Expected Mid-June](https://www.techtimes.com/articles/317328/20260528/grok-ai-new-model-triples-parameter-count-targets-coding-lead-release-expected-mid-june.htm) — Tech Times, 2026-05-28
- [Grok 5: Release Date & All We Know So Far (June 2026)](https://felloai.com/all-we-know-so-far-about-grok-5/) — Fello AI, 2026-06-05
- [Grok 4.3 — models](https://docs.x.ai/developers/models/grok-4.3) — xAI, 2026-06-03