# xAI Grok 4.1 Fast Non-Reasoning — deprecation status

> Yes — Grok 4.1 Fast Non-Reasoning is retired. xAI shut it down on May 15, 2026; API requests to it now fail. The recommended replacement is Grok 4.3.

The straight-through twin of Grok 4.1 Fast Reasoning, same weights and same 2M window with thinking turned off. Splitting one model into two ids by reasoning mode is exactly what grok-4.3's reasoning_effort parameter made unnecessary, and both halves went at once.

## Facts

| Field | Value |
| --- | --- |
| Provider | xAI |
| Model id | `grok-4-1-fast-non-reasoning` |
| Status | Retired |
| Shutdown | May 15, 2026 (80 days ago) |
| Last verified | August 3, 2026 |

## Replacement

- **Grok 4.3** (`grok-4.3`) · recommended — https://modeldeprecations.dev/xai/grok-4.3
  - Maps to grok-4.3 with reasoning effort "none". Set that explicitly rather than relying on the redirect, or a later change to the default leaves you paying for thinking tokens you never asked for.
  - Parameters it accepts: https://modelparams.dev/models/xai/grok-4.3

```diff
- "model": "grok-4-1-fast-non-reasoning"
+ "model": "grok-4.3"
```

## FAQ

### When did xAI Grok 4.1 Fast Non-Reasoning shut down?

xAI Grok 4.1 Fast Non-Reasoning shut down on May 15, 2026. API requests to it fail from that date onward.

### What replaces xAI Grok 4.1 Fast Non-Reasoning?

xAI recommends Grok 4.3 (`grok-4.3`) as the replacement for xAI Grok 4.1 Fast Non-Reasoning. Maps to grok-4.3 with reasoning effort "none". Set that explicitly rather than relying on the redirect, or a later change to the default leaves you paying for thinking tokens you never asked for.

### Is xAI Grok 4.1 Fast Non-Reasoning still available?

No. xAI Grok 4.1 Fast Non-Reasoning is retired and no longer served — it was shut down on May 15, 2026. Calls to it return an error.

### How do I migrate from xAI Grok 4.1 Fast Non-Reasoning to grok-4.3?

Swap the model id in your API call: replace "grok-4-1-fast-non-reasoning" with "grok-4.3". Maps to grok-4.3 with reasoning effort "none". Set that explicitly rather than relying on the redirect, or a later change to the default leaves you paying for thinking tokens you never asked for.

## Sources

- [Grok Model Retirement on May 15, 2026](https://docs.x.ai/developers/migration/may-15-retirement) — accessed 2026-08-03

---

Canonical page: https://modeldeprecations.dev/xai/grok-4-1-fast-non-reasoning
JSON: https://modeldeprecations.dev/api/v1/models/xai/grok-4-1-fast-non-reasoning.json