deepseek-r1 vs grok-4
Side-by-side comparison of deepseek-r1 and grok-4— benchmarks, pricing, context window and capabilities. Both are accessible through Requesty's unified API. grok-4 outperforms deepseek-r1 on 11 of 11 shared benchmarks.
deepseek-r1
Input / 1M
$4.00
Output / 1M
$4.00
Context
64K
Model ID
novita/deepseek/deepseek-r1
xAI Corp.
grok-4
Input / 1M
$3.00
Output / 1M
$15.00
Context
256K
Model ID
xai/grok-4
Benchmark comparison
Intelligence Indexreasoning
deepseek-r127.1%
grok-441.5%
Coding Indexcoding
deepseek-r124.0%
grok-440.5%
Math Indexmath
deepseek-r176.0%
grok-492.7%
GPQA Diamondreasoning
deepseek-r181.3%
grok-487.7%
AIME 2025math
deepseek-r176.0%
grok-492.7%
LiveCodeBenchcoding
deepseek-r177.0%
grok-481.9%
Terminal-Bench Hardagentic
deepseek-r115.9%
grok-437.9%
τ²-Benchagentic
deepseek-r136.5%
grok-474.9%
SciCodecoding
deepseek-r140.3%
grok-445.7%
MMLU Proknowledge
deepseek-r184.9%
grok-486.6%
Humanity's Last Examreasoning
deepseek-r114.9%
grok-423.9%
Scores sourced from official model cards, Artificial Analysis, and public leaderboards. Benchmarks measure specific skills and don't capture every aspect of model quality.
Pricing & specifications
| deepseek-r1 | grok-4 | |
|---|---|---|
| Input price / 1M | $4.00 | $3.00 |
| Output price / 1M | $4.00 | $15.00 |
| Context window | 64K tokens | 256K tokens |
| Max output | — | — |
| Vision input | — | Yes |
| Tool calling | Yes | Yes |
| Reasoning | — | — |
| Prompt caching | — | Yes |
| Computer use | — | Yes |
| Provider | Novita AI | xAI Corp. |
Questions people ask
Is deepseek-r1 better than grok-4?
grok-4 outperforms deepseek-r1 on 11 of 11 shared benchmarks. See the benchmark comparison above for specifics — deepseek-r1 and grok-4 have different strengths across reasoning, coding, math and multimodal tasks.
Which is cheaper — deepseek-r1 or grok-4?
deepseek-r1 is cheaper. deepseek-r1 costs $4.00/$4.00 per 1M input/output tokens, while grok-4 costs $3.00/$15.00.
Can I use deepseek-r1 and grok-4 through the same API?
Yes. Requesty provides a single OpenAI-compatible API that routes to both. Change just the "model" parameter to switch — "novita/deepseek/deepseek-r1" or "xai/grok-4" — no other code changes needed.
What are the context windows?
deepseek-r1 supports up to 64K tokens of context. grok-4 supports up to 256K tokens. Longer context means you can feed larger documents or codebases in a single prompt, though quality often degrades past 128K for most models.
Switch between deepseek-r1 and grok-4 with one line of code
Requesty provides a single OpenAI-compatible API for 400+ models. Change the model parameter, not your code.
Get started free