Grok 5 Release Date: What's Real, What's Musk's Claim, and How to Know First

Ask "when does Grok 5 ship?" and the honest answer is a date x​AI has never given. As of August 13, 2026 — the day after Grok 4.6 landed — the x​AI model list at docs.x.ai still tops out at grok-4.6. There is no grok-5 model ID, no pricing, no context window, no benchmark card, and no Grok app entry. Grok 5 is real and officially in training, but everything else about it is a claim with a person's name attached. This article separates the confirmed from the said-from-a-stage, sets the timeline against the flagship x​AI actually shipped yesterday, and ends with a way to know first when Grok 5 exists.

The one-line answer, if that's all you came for: Grok 5 is not released, and the closest thing to an official statement is Musk saying — most recently on an August 5, 2026 call — that it should arrive within 2026. That is not a date, and x​AI has never published one. The question that actually matters for anyone building today is what to run until then, and the answer is the flagship that exists: Grok 4.6, which shipped August 12 at $2 per million input tokens and $6 per million output.

What is confirmed about Grok 5 (it's a short section)

Everything that is actually confirmed about Grok 5 fits in a couple of lines, because x​AI has been disciplined about saying almost nothing. The company's January 28, 2026 Series E funding announcement confirmed that Grok 5 was in training on the Colossus 2 cluster in Memphis. That is the one statement that reads like a fact. Since then, x​AI's own pages have added nothing: the model list still ends at grok-4.6, the release notes carry no fifth-generation entry, and no price card, context window, or benchmark has been published under the Grok 5 name.

Everything else with a number is a claim — labeled as such

The gap between what's confirmed and what's circulating is where most "Grok 5 specs" coverage lives, so the honest thing is to list the circulating figures and attach the correct sourcing label to each.

• A 2.1-trillion-parameter pretraining — Musk's own description of this generation's new pretraining, per his public remarks. It is his claim, not a confirmed specification, and the same figure has also been attached to the intermediate Grok 4.7 rather than to Grok 5 itself. x​AI has published no parameter count for Grok 5.

• A roughly 6-trillion-parameter flagship — the figure Musk cited at a Baron Capital conference in late 2025, describing the target as a "higher intelligence density per gigabyte" design. Also unverified; it is a founder's number, not a spec-sheet number.

• A 1.5-to-2-million-token context window — widely repeated across release trackers, but sourced to community reports and promotional material rather than to x​AI. Unconfirmed.

• "Grok 5 is AGI" — Musk escalated from a "10% chance" remark at the same Baron event to a flat "Grok 5 is AGI" on X. No test definition, no benchmark, no independent evidence, and no reason to treat it as anything but a product claim.

• Trained on SpaceX's engineering corpus — reported by Business Insider after the August 5 call, with Musk positioning Grok 5 as "the strongest engineer-type AI." Plausible, sourced to Musk's remarks, and still not a release fact.

The timeline, and the pattern you should notice

The only official word on timing is "within 2026." Musk's August 5 call repeated that window and sketched the path to it: Grok 4.6 "next week," Grok 4.7 in three to four weeks, Grok 5 before year-end. That call is where most of the "Grok 5 release date" coverage comes from, and it deserves the same scrutiny x​AI's models get, because the track record of these verbal deadlines is measured. Musk's July framing pointed at Grok 4.6 around August 7; it actually shipped August 12. Grok 5's original target was Q1 2026, then a Q2 window — both missed. Prediction markets that briefly priced a pre-June-30 release near 33 percent now sit in single digits.

None of that means Grok 5 is "late" by some objective standard — there is no schedule to be late against. It means "within 2026" is the honest window, and any date more specific than that is someone's hope, not a plan you should route a roadmap around.

Where Grok 5 sits relative to Grok 4.6

To reason about what Grok 5 must beat, anchor on the model x​AI actually shipped yesterday. Grok 4.6 is the current flagship: $2 per million input and $6 per million output on prompts under 200,000 tokens (doubling to $4/$12 above that threshold), a 500,000-token context window, a February 1, 2026 knowledge cutoff, and an Artificial Analysis Intelligence Index of 61. Grok 4.5, released in July, sits below it. Musk has described Grok 4.7 — expected roughly three to four weeks after the August 5 call — as the intermediate step on the way to Grok 5. In other words, Grok 5 is not the next Grok; Grok 4.7 is. Grok 5 is the flagship beyond that, the "within 2026" capstone of a cadence x​AI has been running at roughly one model every few weeks.

What this means for teams calling Grok today

For developers, the practical effect is that the ceiling on the API is Grok 4.6, and the model you call is exactly that — not the one in the teasers. On OrcaRouter's catalog the Grok series currently runs Grok 4.6 and Grok 4.5, billed at the provider's list price with 0% markup, so the $2/$6 you put in your cost model is the $2/$6 on your invoice. When Grok 5 ships and a provider onboards it, we'll list it the same way, at list price, on the same day — one key, no new integration, no second contract.

That same routing layer is also the low-risk way to move when Grok 5 does land. A brand-new model's scores are vendor claims until independent labs reproduce them — the first-week numbers are precisely the ones to distrust — and automatic failover lets you point traffic at it without betting the production path on a vendor card. If a regression shows up on real requests, the next request falls back to a provider that still answers. That's the difference between trying an unproven model and betting your pipeline on it.

How to know first (better than the rumor mill)

Everything above goes stale the day x​AI flips a switch, so here is the check that beats every tracker:

• Watch the model list, not the tweets. The moment Grok 5 is real, docs.x.ai/docs/models gains a grok-5 model ID with pricing and a context window attached. That is literally how Grok 4.6 appeared — the model list first, the announcement after. Bookmark it and check it weekly.

• x​AI's release notes and its X account are the vendor's own channels, but the model list is the artifact, because it is what the API actually serves. A model ID on the docs page is a fact; a mention on X is a statement.

• Watch the OrcaRouter model catalog. A Grok 5 model page on orcarouter.ai is the "you can call it today" signal — when a provider has onboarded it and we're listing it, it's routable through one API at list price.

• For consumer-side visibility, the Grok app gains the model as an entry point. API availability and app availability can diverge, so which signal you watch depends on whether you're building or just curious.

The state of play

Grok 5 is a real model in active training, and it is not released, has no date, and has no spec sheet. The confirmed record is two sentences long; everything else is a claim that needs a third-party lab to become a fact. If you came here for a release date, the honest answer is "within 2026," with a strong prior, based on how Grok 4.5 and Grok 4.6 actually shipped, that the verbal timeline slips before it lands. The winning move for anyone building today is not to wait — it's to benchmark against what's live at $2/$6 and switch the moment the model list changes. When Grok 5 ships, it should land in your key, not in your calendar.