Leaderboard · Model · Updated
Grok 4
xAI's reasoning model — leading on Humanity's Last Exam at release.
At a glance
| Provider | xAI |
|---|---|
| Released | 2025-07 |
| Tier | frontier |
| License | Closed / proprietary |
| Modalities | text, image |
| Context window | 256k tokens |
| Max output | 64k tokens |
| API price · input | $3.00 / 1M tokens |
| API price · output | $15.00 / 1M tokens |
Benchmark performance
How Grok 4 stacks up against the current leader (GPT-5) and the median model in the leaderboard:
| Benchmark | Grok 4 | GPT-5 | Median |
|---|---|---|---|
| Chatbot Arena Elo | 1378 | 1410 | 1320 |
| MMLU-Pro | 86.6 | 86.8 | 78.0 |
| GPQA Diamond | 87.7 | 87.3 | 65.0 |
| MATH | 95.0 | 96.7 | 78.3 |
| HumanEval | 93.0 | 95.1 | 92.0 |
| SWE-Bench Verified | 72.0 | 74.9 | 49.0 |
Numbers compiled from provider technical reports and Chatbot Arena. See methodology for the composite-score formula.
OpenRouter routes your requests across Grok 4, GPT-5, Claude, Gemini, and 100+ other models behind a single API key — pay-as-you-go, no monthly minimum. Try OpenRouter → (affiliate · supports this site)
What does Grok 4 cost in practice?
API pricing is $3.00 per 1M input tokens and $15.00 per 1M output tokens. Assuming a 50/50 input/output split, here is what that looks like at three workload sizes:
| Volume | Per day | Per month | Per year |
|---|---|---|---|
| 1M tokens/day | $9.00 | $270.00 | $3,285 |
| 10M tokens/day | $90.00 | $2,700 | $32,850 |
| 100M tokens/day | $900.00 | $27,000 | $328,500 |
Strengths & weaknesses
Where it shines
- GPQA: 87.7 (rank #2 of 29, top tier)
- MMLU-Pro: 86.6 (rank #3 of 29, top tier)
- HumanEval: 93.0 (rank #4 of 30, top tier)
Where it lags
No clear weaknesses across published benchmarks.
Best alternatives
The closest models to Grok 4 by tier and benchmark score:
| Model | Score | $ in / out | Context | Action |
|---|---|---|---|---|
| Claude Opus 4.1 Anthropic |
83.6 | $15.00 / $75.00 | 200k | Try → · vs Grok |
| o3 OpenAI |
83.7 | $2.00 / $8.00 | 200k | Try → · vs Grok |
| GPT-5 OpenAI |
86.0 | $1.25 / $10.00 | 400k | Try → · vs Grok |
| Gemini 2.5 Pro |
80.9 | $1.25 / $10.00 | 2M | Try → · vs Grok |
| o1 OpenAI |
75.7 | $15.00 / $60.00 | 200k | Try → · vs Grok |
Frequently asked questions
Is Grok 4 a good model?
Grok 4 scores 83.6 on the llmrank.top composite (rank #3 of 30). It is in the top tier (the leader, GPT-5, scores 86.0). Whether it's the right fit depends on your workload — see the use-case discussion on this page.
How much does Grok 4 cost?
Grok 4 is priced at $3.00 per 1M input tokens and $15.00 per 1M output tokens (USD). For a 10M-token-per-day workload split 50/50, that works out to roughly $90.00/day or $32,850/year.
What is Grok 4's context window?
256k tokens. That covers most multi-document workloads.
Is Grok 4 open source?
No — it is a closed (proprietary) model accessed only via the provider API.
What is Grok 4's SWE-Bench score?
Grok 4 scores 72.0% on SWE-Bench Verified — the benchmark that measures real-world GitHub issue resolution. Its Chatbot Arena Elo is 1378.
What are the best alternatives to Grok 4?
Closest alternatives by tier and score: Claude Opus 4.1, o3, GPT-5. See the alternatives section on this page for side-by-side numbers.
Related: Claude Opus 4.1 · o3 · GPT-5 · Full leaderboard
Spotted out-of-date numbers? Open an issue — corrections usually ship within 24h.
Try Grok 4 now
Direct link to the official playground — or use OpenRouter to A/B test it against any other model on a single API key.