8 Retired Grok Slugs Still Work — Most Bill at grok-4.3 Rates

8 Retired Grok Slugs Still Work — Most Bill at grok-4.3 Rates — illustration

If your code still sends requests to grok-4-1-fast-non-reasoning, grok-3, or six other Grok model IDs, those requests still work. They worked before 15 May 2026 and they have kept working since. Nothing errored, nothing warned, nothing in your logs changed. What changed that day was the bill.

Six of the eight are being served by grok-4.3 and billed at grok-4.3 prices: $1.25 per 1M input tokens and $2.50 per 1M output. The two exceptions are grok-code-fast-1, now served by grok-build-0.1, and grok-imagine-image-pro, now served by xAI’s image models.

Last verified: 7 November 2026 against the migration notices and current model docs.

Updated 7 November 2026: grok-imagine-image-quality, the target for grok-imagine-image-pro, was itself retired on 2 November, so that row now points further down the chain. The pricing table adds grok-4.6, grok-4.7 and a cached-input column. The note on the name now reflects that the company calls itself SpaceXSI.

(xAI was absorbed into SpaceX and now presents itself as SpaceXSI in its own docs. The API itself still says xAI. More on that below.)

The eight slugs

From xAI’s model retirement notice, effective 15 May 2026 at 12:00 PM PT:

Retired slugNow served by
grok-4-1-fast-reasoninggrok-4.3, low reasoning effort
grok-4-1-fast-non-reasoninggrok-4.3, none reasoning effort
grok-4-fast-reasoninggrok-4.3, low reasoning effort
grok-4-fast-non-reasoninggrok-4.3, none reasoning effort
grok-4-0709grok-4.3, low reasoning effort
grok-code-fast-1grok-build-0.1
grok-3grok-4.3, none reasoning effort
grok-imagine-image-progrok-imagine-image-quality, which since 2 November is itself served by grok-imagine-image-2.0 at low quality

The image row got one hop longer this month. grok-imagine-image-quality was retired on 2 November 2026, and xAI’s image migration notice says:

Requests to grok-imagine-image-quality on /v1/images/generations and /v1/images/edits are served by grok-imagine-image-2.0 with quality: “low”.

grok-imagine-image-pro follows the same chain, so both old slugs now end up at grok-imagine-image-2.0 on low quality. This one went the other way on price: the same doc says low quality “is $0.01 per image cheaper than grok-imagine-image-quality at every resolution”, and usage.cost_in_usd_ticks on responses reflects $0.04 per 1K image and $0.06 per 2K image, plus $0.01 per input image on edits. If you’re grepping for retired image slugs, add grok-imagine-image-quality to the pattern below.

A note on the name

The company that retired these models was xAI. It isn’t called that any more.

SpaceX acquired xAI in a merger that closed 2 February 2026, Musk said in May that xAI would be dissolved as a separate company, and on 6 July 2026 the rebrand to SpaceXAI was made official with a new logo and a renamed X account. That name didn’t stick either. The company now presents itself as SpaceXSI (SpaceX Super Intelligence) in its own docs.

When I first checked on 7 August 2026, the docs at docs.x.ai said “the xAI API” and made no mention of SpaceX or SpaceXAI on the retirement notice. That’s changed. As of 7 November the docs root says “Available on the SpaceXSI API”, and the release notes use the SpaceXSI name throughout, down to “SpaceXSI’s frontier model” and the “SpaceXSI console”.

The parts you actually code against haven’t moved. The base URL in the examples is still https://api.x.ai/v1, the key placeholder is still YOUR_XAI_API_KEY, and bug reports still go to [email protected]. The models are still Grok.

So: the retirement was announced by xAI, the company now calls itself SpaceXSI, and the thing you actually type into your config didn’t change at all. This page says “xAI” where it’s quoting the migration docs, because that’s what they say.

Why this one is easy to miss

Most deprecations announce themselves. You get a 404, a deprecation header, something in a changelog you eventually read. This one is the opposite, and xAI says so plainly:

After May 15, 2026 at 12:00 PM PT, requests to the retired model slugs above will automatically redirect to grok-4.3. The slugs themselves continue to resolve, so you do not need to change your code to avoid breakage.

That’s genuinely considerate engineering — nobody’s production traffic fell over. But the same doc adds:

If you continue sending requests to a deprecated slug after May 15, please be aware that you will be billed at grok-4.3 pricing

So the failure mode isn’t an outage. It’s a bill that changed while every string in your codebase stayed the same. If you set grok-4-1-fast-non-reasoning in an env var eight months ago and haven’t looked since, you are on grok-4.3 pricing and your config still says otherwise. The one text slug that bills differently is grok-code-fast-1: the grok-build-0.1 model page lists it as an alias (“Aliases: grok-code-fast-1, grok-code-fast, grok-code-fast-1-0825”), so it bills as grok-build-0.1 even though the retirement notice’s billing line speaks of deprecated slugs in general.

What you actually pay now

Current xAI model pricing, re-checked 7 November 2026:

ModelInput / 1MCached input / 1MOutput / 1M
grok-4.7 (<200k ctx)$2.00$0.50$6.00
grok-4.7 (≥200k ctx)$4.00$1.00$12.00
grok-4.6 (<200k ctx)$2.00$0.50$6.00
grok-4.6 (≥200k ctx)$4.00$1.00$12.00
grok-4.5 (<200k ctx)$2.00$0.30$6.00
grok-4.5 (≥200k ctx)$4.00$0.60$12.00
grok-4.3 (<200k ctx)$1.25$0.20$2.50
grok-4.3 (≥200k ctx)$2.50$0.40$5.00
grok-4.20 reasoning / non-reasoning (<200k ctx)$1.25$0.20$2.50
grok-4.20-multi-agent (<200k ctx)$1.25$0.20$2.50
grok-build-0.1 (<200k ctx)$1.00$0.20$2.00

grok-4.6 and grok-4.7 match grok-4.5 on input and output, but cached input costs more: $0.50 against $0.30 under 200k.

Published Grok API prices under 200k context, and where retired slugs bill Horizontal paired bar chart of xAI's published per-million-token prices for context under 200k, verified 7 November 2026: grok-4.5 costs 2 dollars input and 6 dollars output; grok-4.3 costs 1.25 input and 2.50 output and is highlighted because six of the eight retired Grok slugs, all but grok-code-fast-1 and grok-imagine-image-pro, are served by it and bill at its rates; grok-4.20 reasoning and non-reasoning cost 1.25 and 2.50; grok-4.20-multi-agent costs 1.25 and 2.50; grok-build-0.1 costs 1 dollar input and 2 dollars output. The point: six of the eight retired slugs bill at grok-4.3's 1.25 and 2.50 regardless of what the model name in your config suggests. Source: docs.x.ai migration and models pages. Where a retired Grok slug actually bills Published $ per 1M tokens, context under 200k · 6 of 8 retired slugs bill as grok-4.3 Input Output $0 $2 $4 $6 grok-4.5 $2.00 $6.00 grok-4.3 6 of 8 retired slugs bill here $1.25 $2.50 grok-4.20 reasoning / non-reasoning $1.25 $2.50 grok-4.20-multi-agent $1.25 $2.50 grok-build-0.1 $1.00 $2.00 grok-4.5 and grok-4.3 double at 200k+ context. Verified 7 Nov 2026 against docs.x.ai; slugs and redirects from xAI's migration doc.
xAI's published under-200k prices per 1M tokens. Six of the eight retired slugs, all but grok-code-fast-1 and grok-imagine-image-pro, are served by grok-4.3 and bill at its $1.25 / $2.50, regardless of the model name your config still shows. Selected models. Verified 7 November 2026 against xAI's own docs.

A note on the old prices. You’ll see figures circulating for what the retired fast models used to cost. We’re not stating any of them as fact: xAI no longer publishes pricing for retired models and we couldn’t reach an archived copy to confirm it. For your own delta, the only record of what you were actually charged is your xAI billing history either side of 15 May.

You don’t need that number to see the problem, though. Everything above is published, and it already shows you’re on the mid tier: grok-4.3 at $1.25/$2.50 when grok-build-0.1 sits right there at $1.00/$2.00. You didn’t choose that — a redirect did.

What isn’t in doubt is the direction for the four fast slugs. grok-4-1-fast and grok-4-fast, reasoning and non-reasoning, were the cheap, fast tier, and grok-4.3 is not. grok-3 and grok-4-0709 were flagship models, so if you were on one of those, the move may have gone the other way.

Check in thirty seconds

grep -rEn 'grok-4-1-fast|grok-4-fast|grok-4-0709|grok-code-fast-1|grok-3\b|grok-imagine-image-pro|grok-imagine-image-quality' \
  --include='*.js'   --include='*.ts'   --include='*.py'  --include='*.go'  \
  --include='*.rb'   --include='*.rs'   --include='*.java' --include='*.json' \
  --include='*.yaml' --include='*.yml'  --include='*.env' --include='*.toml' .

(One --include per extension on purpose. The tidier-looking --include='*.{js,ts,py}' matches nothing at all: the braces are inside quotes so the shell never expands them, and --include doesn’t do brace alternation — so it fails silently and looks like a clean bill of health.)

Also check the places code search misses: environment variables in your hosting dashboard, CI secrets, serverless config, and any model name stored in a database or feature flag.

If you get hits, you have a decision rather than an emergency. Requests still succeed. But you should pick the model deliberately instead of inheriting one, and grok-build-0.1 at $1.00/$2.00 is cheaper than grok-4.3 if the work suits it.

The wider lesson

This is the same class of problem as free tiers that quietly shrink — the number you built against stops being the number you’re on. Silent redirect is becoming the normal way model providers retire things, because breaking production is worse press than a quiet price change. It’s the right call for uptime. It does mean the old assumption — if it still runs, nothing changed — no longer holds for AI APIs.

The practical habit: treat model IDs like dependency versions. Pin them somewhere visible, and re-read the provider’s model page on a schedule rather than when something breaks. Nothing will break.

Keep reading

Frequently Asked Questions

Which Grok model slugs were retired on 15 May 2026?

Eight: grok-4-1-fast-reasoning, grok-4-1-fast-non-reasoning, grok-4-fast-reasoning, grok-4-fast-non-reasoning, grok-4-0709, grok-code-fast-1, grok-3, and grok-imagine-image-pro.

Do retired Grok model IDs return an error?

No. xAI's migration doc states the slugs continue to resolve and requests redirect automatically, so nothing in your code breaks and nothing signals that anything changed.

What do retired Grok slugs cost now?

Six of the eight bill at grok-4.3 pricing: $1.25 per 1M input tokens and $2.50 per 1M output under 200k context, or $2.50/$5.00 at 200k and above. The two exceptions are grok-code-fast-1, served by grok-build-0.1 ($1.00/$2.00 under 200k), and grok-imagine-image-pro, which now ends up at grok-imagine-image-2.0 on low quality and is priced per image.

Is xAI now SpaceXSI?

Yes, as far as its own docs go. SpaceX absorbed xAI in a merger that closed on 2 February 2026, the company became SpaceXAI in July, and docs.x.ai now presents it as SpaceXSI (SpaceX Super Intelligence), though the API base URL (api.x.ai), the key placeholder (YOUR_XAI_API_KEY) and the support address ([email protected]) still say xAI.

How do I check whether I'm affected?

Grep your codebase and environment config for the eight retired slugs. If any of the six text and reasoning slugs appear, you are being billed at grok-4.3 rates regardless of what the model name suggests. grok-code-fast-1 is served by grok-build-0.1, and grok-imagine-image-pro by xAI's image models.

Written by Hirak Banerjee

Indie dev and maker. I build AI-powered apps and write about the tools I actually use. Follow on X · GitHub

Get told when these numbers change

Every figure here is re-checked against the vendor's own docs. Leave your email and we'll tell you when one moves — only when something actually changes.

Join builders who ship faster. No spam.

Comments