DeepSeek V4 Pro 0813 vs Grok 4.6

At a glance

• Capability index

  • DeepSeek V4 Pro 0813155.3
  • Grok 4.6156.4

  Grok 4.6 is 1.1 points higher.

• Price per 1M tokens

  • DeepSeek V4 Pro 0813$1.98
  • Grok 4.6$3.00

  DeepSeek V4 Pro 0813 is 34% cheaper.

• Context window

  • DeepSeek V4 Pro 08131M tokens
  • Grok 4.6500K tokens

  DeepSeek V4 Pro 0813 takes in 2.1 times as much.

• Released

  • DeepSeek V4 Pro 0813Aug 12, 2026
  • Grok 4.6Aug 12, 2026

  Released on the same day.

Where each scores higher

On the 14 tests both have taken, each on its own scale. A gap of a point or two can sit within a test’s margin of error.

9 more tests have results for only one of them; the table below lists every result.

Everything side by side

Comparison of DeepSeek V4 Pro 0813 vs Grok 4.6

Measure	DeepSeek V4 Pro 0813DeepSeek	Grok 4.6xAI
Model
Maker	DeepSeek	xAI
Sold here by	CrofAI	xAI
API model ID	deepseek-v4-pro-0813	grok-4.6
Released	Aug 12, 2026	Aug 12, 2026
Knowledge cutoff	Not reported	Feb 1, 2026
Weights	Open: downloadable	Closed: hosted access only
Developer’s country	China	United States
Hosts selling it	42	35including xAI directly
Training compute	9.7 × 10²⁴ FLOPEpoch AI estimate	Not published
Test results
Capability index	155.3#29 of 274	156.4 (best of these models)#21 of 274
LiveBench overall	77.4%#23 of 66	78.0% (best of these models)#18 of 66
Arena text	Not tested	1454#73 of 413
LiveBench coding	77.2% (best of these models)#38 of 66	76.8%#42 of 66
Arena WebDev	Not tested	1620#23 of 138
Arena coding	Not tested	1507#64 of 408
LiveBench mathematics	95.1% (best of these models)#16 of 66	92.6%#25 of 66
Mock AIME 2024–2025	98.6%#20 of 297	99.2% (best of these models)#15 of 297
FrontierMath Tiers 1–3	64.6%#32 of 114	66.0% (best of these models)#28 of 114
FrontierMath Tier 4	26.8%#39 of 70	31.7% (best of these models)#33 of 70
Arena math	Not tested	1448#78 of 396
SimpleQA Verified	52.9% (best of these models)#21 of 86	49.3%#27 of 86
LiveBench agentic coding	55.0%#28 of 66	57.0% (best of these models)#22 of 66
DeepSWE	Not tested	67.5%#20 of 69
CursorBench	Not tested	41.4%#24 of 62
LiveBench language	82.1%#26 of 66	83.7% (best of these models)#19 of 66
LiveBench instructions	67.7%#36 of 66	71.9% (best of these models)#20 of 66
Arena creative writing	Not tested	1445#52 of 411
Arena instruction following	Not tested	1451#63 of 413
LiveBench data analysis	79.2% (best of these models)#18 of 66	73.9%#40 of 66
LiveBench reasoning	85.8%#34 of 66	90.5% (best of these models)#12 of 66
GPQA Diamond	91.7%#29 of 319	94.0% (best of these models)#11 of 319
Chess puzzles	47.0% (best of these models)#14 of 227	40.0%#24 of 227
Arena hard prompts	Not tested	1479#67 of 413
Price per million tokens
Input	$0.35 (best of these models)	$2.00
Output	$0.80 (best of these models)	$6.00
Blended, 3 input to 1 output	$0.462 (best of these models)	$3.00
Cached input	$0.01 (best of these models)	$0.50
Long requests	No separate price reported	$4.00 in, $12.00 outabove 200K tokens
The maker’s own price	Not sold directly	This listing
Limits
Context window	1M tokens (best of these models)	500K tokens
Max input	1M tokens (best of these models)	372K tokens
Max output	131.1K tokens	500K tokens (best of these models)
Features
Reads	text	images, PDFs and text
Produces	text	text
Reasoning	Yeseffort none, low, high, max; can be switched off; thinking budget can be set	Yeseffort low, medium, high, xhigh
Tool calling	Yes	Yes
Structured output	Yes, such as JSON	Yes, such as JSON
File attachments	No	Yes
Temperature setting	Supported	Supported

Bold marks the better value in each row. Prices are each model’s own list price, or the middle price across hosts when the maker does not sell it, blended as three input tokens for every output token. Test results as published by LiveBench (Apache-2.0), Epoch AI (CC BY 4.0), Datacurve DeepSWE, compiled by Epoch AI (CC BY 4.0), Cursor CursorBench, compiled by Epoch AI (CC BY 4.0), Arena (CC BY 4.0); model details and prices from models.dev (MIT). Confirm pricing with the provider before use.