Grok-3 vs Grok 4
Grok-3 (2025) and Grok 4 (2025) are frontier-tier reasoning models from xAI. Grok-3 ships a 1M-token context window, while Grok 4 ships a 256k-token context window. On MMLU PRO, Grok 4 leads by 7.1 pts. On pricing, Grok-3 costs $0.8/1M input tokens versus $1.25/1M for the alternative. This comparison covers specs, pricing, capabilities, benchmarks, provider availability, and production fit. It focuses on practical selection signals rather than broad model-family marketing.
Grok-3 is ~56% cheaper at $0.8/1M; pay for Grok 4 only for coding workflow support.
Decision scorecard
Local evidence first| Signal | Grok-3 | Grok 4 |
|---|---|---|
| Decision fit | Coding, RAG, and Agents | Coding, RAG, and Agents |
| Context window | 1M | 256k |
| Cheapest output | $2.4/1M tokens | $2.5/1M tokens |
| Provider routes | 4 tracked | 4 tracked |
| Shared benchmarks | 2 rows | MMLU PRO leader |
Decision tradeoffs
- Grok-3 has the larger context window for long prompts, retrieval packs, or transcript analysis.
- Grok-3 has the lower cheapest tracked output price at $2.4/1M tokens.
- Local decision data tags Grok-3 for Coding, RAG, and Agents.
- Grok 4 leads the largest shared benchmark signal on MMLU PRO by 7.1 points.
- Grok 4 uniquely exposes Code execution in local model data.
- Local decision data tags Grok 4 for Coding, RAG, and Agents.
Monthly cost at traffic
Estimate token spend from the cheapest tracked input and output prices on this page.
Grok-3
$1,240
Cheapest tracked route: Chutes AI
Grok 4
$1,625
Cheapest tracked route: xAI Console
Estimated monthly gap: $385. Batch, cache, and negotiated pricing are excluded from this local estimate.
Switch friction
- Provider overlap exists on Microsoft Foundry, OpenRouter, and xAI Console; start route-level A/B tests there.
- Grok 4 is $0.1/1M tokens higher on cheapest tracked output pricing, so quality gains need to justify the spend.
- Grok 4 adds Code execution in local capability data.
- Provider overlap exists on OpenRouter, xAI Console, and Microsoft Foundry; start route-level A/B tests there.
- Grok-3 is $0.1/1M tokens lower on cheapest tracked output pricing before cache, batch, or negotiated discounts.
- Check replacement coverage for Code execution before moving production traffic.
Specs
Pricing and availability
| Pricing attribute | Grok-3 | Grok 4 |
|---|---|---|
| Input price | $0.8/1M tokens | $1.25/1M tokens |
| Output price | $2.4/1M tokens | $2.5/1M tokens |
| Providers |
Capabilities
| Capability | Grok-3 | Grok 4 |
|---|---|---|
| Vision | Yes | Yes |
| Multimodal | Yes | Yes |
| Reasoning | Yes | Yes |
| Function calling | Yes | Yes |
| Tool use | Yes | Yes |
| Structured outputs | Yes | Yes |
| Code execution | No | Yes |
Benchmarks
| Benchmark | Grok-3 | Grok 4 |
|---|---|---|
| MMLU PRO | 79.9 | 87.0 |
| Aider Polyglot | 53.3 | 79.6 |
Deep dive
On shared benchmark coverage, MMLU PRO has Grok-3 at 79.9 and Grok 4 at 87, with Grok 4 ahead by 7.1 points; Aider Polyglot has Grok-3 at 53.3 and Grok 4 at 79.6, with Grok 4 ahead by 26.3 points. The largest visible gap is 26.3 points on Aider Polyglot, which matters most when that benchmark mirrors your workload. Treat isolated benchmark wins as directional, because provider routing, prompt style, and tool access can move real application results.
The capability footprint differs most on code execution: Grok 4. Both models share vision, multimodal input, reasoning mode, and function calling, so the practical split is not just feature count. Use those differences to decide whether the page is about raw model quality, agentic coding support, multimodal ingestion, or predictable structured API behavior.
For cost, Grok-3 lists $0.8/1M input and $2.4/1M output tokens, while Grok 4 lists $1.25/1M input and $2.5/1M output tokens on the cheapest tracked provider. A 70/30 input-output blend puts Grok-3 lower by about $0.35 per million blended tokens. Availability is 4 providers versus 4, so concentration risk also matters.
Choose Grok-3 when long-context analysis, larger context windows, and lower input-token cost are central to the workload. Choose Grok 4 when coding workflow support are more important. For production, rerun your own prompts through the exact provider, region, and tool stack you plan to ship.
FAQ
Which has a larger context window, Grok-3 or Grok 4?
Grok-3 supports 1M tokens, while Grok 4 supports 256k tokens. That gap matters most for long documents, large codebases, retrieval-heavy agents, and conversations where earlier context must remain visible. Use this as a quick comparison signal, then confirm the provider-specific limits before committing to production.
Which is cheaper, Grok-3 or Grok 4?
Grok-3 is cheaper on tracked token pricing. Grok-3 costs $0.8/1M input and $2.4/1M output tokens. Grok 4 costs $1.25/1M input and $2.5/1M output tokens. Provider discounts or batch pricing can still change the final bill.
Is Grok-3 or Grok 4 open source?
Grok-3 is listed under Proprietary. Grok 4 is listed under Proprietary. License labels affect whether you can self-host, redistribute weights, or rely only on hosted APIs, so confirm the upstream license before deployment.
Which is better for vision, Grok-3 or Grok 4?
Both Grok-3 and Grok 4 expose vision. The better choice depends on benchmark fit, context budget, pricing, and whether your provider route exposes the same capability surface. Use this as a quick comparison signal, then confirm the provider-specific limits before committing to production.
Which is better for multimodal input, Grok-3 or Grok 4?
Both Grok-3 and Grok 4 expose multimodal input. The better choice depends on benchmark fit, context budget, pricing, and whether your provider route exposes the same capability surface. Use this as a quick comparison signal, then confirm the provider-specific limits before committing to production.
Where can I run Grok-3 and Grok 4?
Grok-3 is available on OpenRouter, xAI Console, Chutes AI, and Microsoft Foundry. Grok 4 is available on Microsoft Foundry, OpenRouter, Replicate API, and xAI Console. Provider coverage can affect latency, region availability, compliance posture, and fallback options.
Continue comparing
Last reviewed: 2026-05-16. Data sourced from public model cards and provider documentation.