← Models

Grok 4

xAI · Grok

Deprecated Grok 4 API route whose old slug now redirects to Grok 4.3 after the May 15, 2026 retirement.

Part of Grok family · Other versions: Grok 4.5 , Grok Build 0.1 , Grok 4.3 , Grok 4.20 , Grok 4.1 Fast , Grok Code Fast 1
Type
language
Context
256K tokens
Max Output
N/A
Status
deprecated
Input
$3/1M tok
Output
$15/1M tok
API Access
Yes
License
proprietary
reasoning tool-use agentic coding long-context real-time-data deprecated
Released July 2025 · Updated July 10, 2026

Overview

Freshness note: Model capabilities, limits, and pricing can change quickly. This profile is a point-in-time snapshot last verified on July 10, 2026.

Grok 4 is now an older Grok reasoning route, not the forward-looking default. xAI’s May 2026 model-retirement notice says older Grok 4 API variants, including grok-4-0709, retired on May 15, 2026 at 12:00 PM PT. Requests to the deprecated slug now redirect to Grok 4.3 with low reasoning effort and are billed at Grok 4.3 pricing.

This page is retained for teams auditing older integrations, pinned behavior, or historical comparisons. Do not route new Grok workloads here unless you are deliberately testing redirect behavior from an old slug.

Capabilities

Historically, Grok 4 was strongest on difficult reasoning, coding, tool-oriented assistants, and long-context analysis where teams wanted a Grok-native flagship. The retirement redirect still targets Grok 4.3, but new difficult coding and agentic workloads should also evaluate Grok 4.5. Grok Build 0.1 remains the lower-cost migration route for old Grok coding configurations.

Technical Details

xAI’s earlier docs listed Grok 4 with a 256K context window and positioned it as a reasoning model rather than a configurable non-reasoning chat tier. If you still have request bodies pinned to Grok 4, test the same prompts and tools against Grok 4.3 explicitly and remove assumptions tied to old unsupported parameters.

xAI’s current retirement and replacement documentation does not publish a separate maximum output-token limit for this deprecated route. Here, maxOutput: 0 means “not separately published,” not zero output capacity. Teams should still set explicit application-level output caps and timeout budgets.

Pricing & Access

Listed API pricing (per 1M tokens):

  • Input: $3.00
  • Output: $15.00

Access options:

  • xAI API platform
  • Grok product surfaces depending on plan/tier
  • partner and enterprise integrations where enabled

Because the deprecated slug redirects to Grok 4.3 pricing after May 15, Grok 4 should be treated as a migration concern rather than a production planning target.

Best Use Cases

Use this page for:

  • auditing older Grok 4 deployments
  • comparing migration outputs against Grok 4.3
  • documenting historical premium Grok behavior
  • cleaning up old model names after the May 15, 2026 redirect change

Avoid starting new Grok 4 integrations. Use Grok 4.5 for the current flagship route, Grok 4.3 for a lower-cost 1M-context alternative, or Grok Build 0.1 for the pinned lower-cost coding API path.

Comparisons

  • Grok 4.3 (xAI): Recommended replacement for new API work and the target migration route for retired Grok 4 variants.
  • Grok 4.5 (xAI): Current flagship for difficult coding and agentic knowledge work.
  • Grok Build 0.1 (xAI): Dedicated coding-agent route for teams moving away from older Grok coding models.
  • Grok 4.20 (xAI): 1M-context preview lane for teams explicitly testing xAI’s beta research direction.
  • Claude Opus 4.7 (Anthropic): Stronger governance and conservative deployment story, while Grok stays more experimental and xAI-native.