Grok 4.3
Grok 4.3 is a xAI reasoning model with an improved architecture and a December 2025 knowledge cutoff, matching the scale of Grok 4.20 within a context window of 1M tokens.
import { streamText } from 'ai'
const result = streamText({ model: 'xai/grok-4.3', prompt: 'Why is the sky blue?'})Playground
Try out Grok 4.3 by xAI. Usage is billed to your team at API rates. Free users (those who haven't made a payment) get $5 of credits every 30 days.
Providers
Route requests across multiple providers. Copy a provider slug to set your preference. Visit the docs for more info. Using a provider means you agree to their terms, listed under Legal.
P50 throughput on live AI Gateway traffic, in tokens per second (TPS). Visit the docs for more info.
P50 time to first token (TTFT) on live AI Gateway traffic, in milliseconds. View the docs for more info.
Direct request success rate on AI Gateway and per-provider. Visit the docs for more info.
More models by xAI
| Model |
|---|
| 256K | 0.3s | 214tps | $1.00/M | $2.00/M | Read: $0.2/M Write: — | $5/K + input costs | — |  | 05/20/2026 | ||||
| 2M | 0.5s | 99tps | $1.25/M | $2.50/M | Read: $0.2/M Write: — | $5/K + input costs | — | | 03/11/2026 | ||||
| 2M | 0.5s | 80tps | $1.25/M | $2.50/M | Read: $0.2/M Write: — | $5/K + input costs | — |  | 03/09/2026 | ||||
| 2M | 2.3s | 1203tps | $1.25/M | $2.50/M | Read: $0.2/M Write: — | $5/K + input costs | — | | 03/09/2026 | ||||
| 1M | 1.3s | 261tps | $0.20/M | $0.50/M | Read:$0.05/M Write:— | — | — | | 07/09/2025 | ||||
| 1M | 0.3s | 72tps | $0.20/M | $0.50/M | Read:$0.05/M Write:— | — | — | | 07/09/2025 |
About Grok 4.3
Grok 4.3 was released April 30, 2026 as xAI's next reasoning model in the Grok family. It matches the scale of Grok 4.20 while refining the underlying architecture, and ships with a December 2025 knowledge cutoff so factual recall is closer to current events than earlier Grok generations.
The model targets reasoning-heavy workloads: mathematical problem solving, code analysis, multi-step planning, and tasks where structured thinking improves answer quality. It supports tool use, vision input, file input, and web search, which makes it suitable for agentic systems that combine reasoning with external context. Implicit caching helps keep costs predictable on repeated prefixes within the context window of 1M tokens.
Grok 4.3 produces up to 1M tokens per response, which leaves room for long reasoning traces alongside the final answer. See the pricing panel on this page for current input and output rates.
What To Consider When Choosing a Provider
- Configuration: Grok 4.3 produces extended reasoning output, which increases token consumption per response. Budget output tokens generously and plan for the reasoning overhead when estimating cost.
- Configuration: The December 2025 knowledge cutoff is newer than earlier Grok models, but Grok 4.3 still benefits from web search or retrieval augmentation for facts that change after that date. See live metrics on this page for current throughput.
- Zero Data Retention: AI Gateway does not currently support Zero Data Retention for this model. See the documentation for models that support ZDR.
- Authentication: AI Gateway authenticates requests using an API key or OIDC token. You do not need to manage provider credentials directly.
When to Use Grok 4.3
Best For
- Reasoning-Heavy Workloads: Mathematical problem solving, multi-step planning, and analytical tasks that benefit from chain-of-thought
- Long-Context Analysis: Document synthesis, codebase review, and research tasks that fit inside 1M tokens of context
- Tool-Using Agents: Workflows that combine reasoning with web search, file input, and external tool calls
- Vision And Multimodal Tasks: Pipelines that mix text reasoning with image understanding
- Recent-Knowledge Questions: Tasks where the December 2025 knowledge cutoff reduces the need for retrieval
Consider Alternatives When
- Latency-Sensitive Endpoints: Grok 4 Fast variants return direct answers with lower reasoning overhead
- High-Volume Classification: Smaller Grok models like Grok 3 Mini Fast cost less per token for simple tasks
- Coding-Specific Pipelines: Grok Code Fast 1 is tuned for fast code generation at lower cost
- Multi-Agent Orchestration: The Grok 4.20 Multi-Agent variant is purpose-built for agent collaboration patterns
Conclusion
Grok 4.3 is xAI's refined reasoning model at the Grok 4.20 scale, with a December 2025 knowledge cutoff and support for tool use, vision, and web search. Teams that need deep reasoning across long contexts and recent world knowledge should evaluate Grok 4.3 against the Grok 4 Fast variants for cost and latency tradeoffs.
Frequently Asked Questions
What is the knowledge cutoff for Grok 4.3?
Grok 4.3 has a December 2025 knowledge cutoff. Pair it with web search or retrieval for facts that change after that date.
How does Grok 4.3 compare to Grok 4.20?
Grok 4.3 matches the scale of Grok 4.20 and refines the underlying architecture. It also carries a newer December 2025 knowledge cutoff.
What is the context window for Grok 4.3?
The context window is 1M tokens, with up to 1M tokens per response.
What capabilities does Grok 4.3 support?
Grok 4.3 supports reasoning, tool use, implicit caching, file input, vision, and web search through AI Gateway.
What does Grok 4.3 cost?
See the pricing section on this page for today's rates. AI Gateway exposes each provider's pricing for Grok 4.3.
How do I authenticate with Grok 4.3 through Vercel AI Gateway?
Use your Vercel AI Gateway API key with
xai/grok-4.3as the model identifier. AI Gateway handles provider routing automatically.Does Vercel AI Gateway support Zero Data Retention for Grok 4.3?
Zero Data Retention is not currently available for this model. Zero Data Retention is offered on a per-provider basis. See https://vercel.com/docs/ai-gateway/capabilities/zdr for details.