Grok 4.3
xAI's May 2026 flagship. 1M context, vision, always-on reasoning, real-time X/web retrieval via DeepSearch.
Grok 4.3 is multimodal AI model from xAI, priced at β¬0.001 per 1M input tokens with a 1M tokens context window.
About this model
0.7
Pricing
API Integration
Use our OpenAI-compatible API to integrate Grok 4.3 into your application.
npm install railwailimport railwail from "railwail";
const rw = railwail("YOUR_API_KEY");
// Simple β just pass a string
const reply = await rw.run("grok-4-3", "Hello! What can you do?");
console.log(reply);
// With message history
const reply2 = await rw.run("grok-4-3", [
{ role: "system", content: "You are a helpful assistant." },
{ role: "user", content: "Explain quantum computing simply." },
]);
console.log(reply2);
// Full response with usage info
const res = await rw.chat("grok-4-3", [
{ role: "user", content: "Hello!" },
], { temperature: 0.7, max_tokens: 500 });
console.log(res.choices[0].message.content);
console.log(res.usage);Deep dive β xAI's Grok 4.3
xAI was founded in March 2023 by Elon Musk together with researchers from DeepMind, OpenAI, Google Research, Microsoft Research and Tesla including Igor Babuschkin and Tony Wu. The company's stated mission is to build maximally truth-seeking AI. Its flagship product is Grok, an LLM-based assistant deeply integrated into X (formerly Twitter) and available via a standalone app and API. Grok 1 weights were released under Apache 2.0 in November 2023 (314B-parameter MoE), Grok 1.5 added 128K context and vision in early 2024, Grok 2 shipped in mid-2024, Grok 3 launched in February 2025 trained on the Colossus supercluster, Grok 4 in mid-2025, and the Grok 4.x line in 2025-2026 culminating in Grok 4.3 on May 1, 2026. xAI raised over $20 billion across multiple rounds with backers including Andreessen Horowitz, Sequoia, Valor Equity, Lightspeed and Saudi funds. The company operates the Colossus cluster in Memphis, the world's largest dedicated AI training site.
Visit xAI βGrok 4.3 was released May 1, 2026 as xAI's frontier model. It is the direct successor to Grok 4 (mid-2025) and Grok 4.20 (early 2026) and follows the same training recipe at significantly larger scale: pretraining on the expanded Colossus supercluster in Memphis (now exceeding 300,000 GPUs), with a multi-trillion-token mixture of web text, code, scientific papers, math content and the continuously growing feed of public posts on the X platform. Post-training relies heavily on large-scale reinforcement learning against verifiable rewards for math, coding and tool-use trajectories. Grok 4.3 ships with always-on reasoning (test-time thinking is no longer optional - the model always reasons before answering) and an upgraded DeepSearch agent that performs multi-step web and X-platform retrieval. The context window jumps to 1M tokens. xAI positions Grok 4.3 as competitive with GPT-5.4 and Gemini 3.1 Pro on math and science benchmarks, with the differentiators being real-time access to X data, a lower refusal rate, and the most generous free-tier API credits of any major provider ($175/month).
- Parameters
- Undisclosed (estimated multi-hundred billion parameters, likely sparse MoE)
- Context
- 1M tokens
- 1M token context window with vision input
- Always-on test-time reasoning
- Upgraded DeepSearch agent for X + web retrieval
- Trained on 300K+ GPU Colossus cluster
- Strong AIME, GPQA, ARC-AGI and SWE-bench scores
- Vision input for images and screenshots
- Function calling and structured JSON output
- Native integration with X timeline and DMs
- Available via xAI API, X Premium+ and the Grok mobile app
- Real-time fresh data through built-in retrieval
- $175/month free API credits via data-sharing programme
- Best for: agentic research, real-time question answering, math/code reasoning, X-platform assistants.
Multi-trillion-token pretraining mixture of web text, code, math, scientific papers and X public posts. Knowledge cutoff updated continuously through retrieval. Post-training applies RLHF and large-scale reinforcement learning against verifiable rewards on math, code and tool-use tasks.
License: Proprietary commercial license via xAI API, X Premium+ and the Grok app. Weights not publicly released for Grok 2 onwards.
Known limitations
- Lower refusal rate increases risk of harmful or biased output
- Limited published third-party safety evaluations
- Real-time retrieval depends heavily on the X corpus
- Regional availability constrained by X service availability
- Weights not released; reproducibility limited
Frequently asked questions
Related Models
View all MultimodalClaude Opus 4.7
Anthropic's April 2026 flagship. 87.6% on SWE-bench Verified, 3x higher image resolution, output self-verification, vision + reasoning.
Claude Sonnet 4.6
Anthropic's balanced mid-tier model from February 2026. Best price/performance for production workloads: 5x cheaper than Opus, near-flagship quality.
Depth Anything v2
Monocular depth-estimation model trained on 595k labeled and 62M unlabeled images. Strong zero-shot generalization in indoor and outdoor scenes.
GPT-5.4
OpenAI's unified flagship combining GPT and o-series reasoning into one model. 1M context, multimodal, top SWE-Bench Pro and OSWorld scores.
Start using Grok 4.3 today
Get started with free credits. No credit card required. Access Grok 4.3 and 100+ other models through a single API.