What are Grok 4.7's benchmark scores?

As of 6 Nov 2026, Grok 4.7 (xhigh effort) scores 58.3 ±2.7 on the BenchLeader Index, strongest in long context (63) and weakest in multimodal (56). That rests on independent benchmark results only, and every score on this page links to its publisher.

How does Grok 4.7 rank?

Grok 4.7 (xhigh effort) is #125 of 587 ranked configurations on 6 Nov 2026. 217 other configurations score within its margin, so the exact position matters less than the band. The leader is Claude Opus 5.5 (max effort) at 69.3.

Is Grok 4.7 better than DeepSeek V4.1 Flash 1017?

As of 6 Nov 2026, DeepSeek V4.1 Flash 1017 (max) and Grok 4.7 (xhigh) are neck and neck on quality (59.2 vs 58.3 on the BenchLeader Index, inside the ±4.6 margin of error), and DeepSeek V4.1 Flash 1017 (max) costs 5.7× less, which makes it the better buy for most workloads. The head-to-head page has the benchmark-by-benchmark detail.

When was Grok 4.7 released?

SpaceXAI released Grok 4.7 on 21 Sep 2026. BenchLeader has results for it at 5 reasoning-effort settings.

SpaceXAIReasoning modelCompare with another model

Grok 4.7

Reasoning effort

Quality#62 of 330

58.3Index

Top quarter

+8.3 on the median

Price#182 of 247

$3.00per 1M tokens

Dearer than most

3.5× the median

Speed#121 of 214

71tokens/s

Below the median

0.9× the median

21 cheaper ranked models score higher: GPT-6.1 Terra (65.3, $2.00), Qwen3.6 Max (62.4, $2.92), Gemini 3.7 Flash (62.2, $1.50) and 18 more.

As of 6 Nov 2026, Grok 4.7 by SpaceXAI scores 58.3 on the BenchLeader Index (587 ranked configurations), with its best results in long context (63) and weakest in multimodal (56). API pricing: $2.00 per 1M input tokens and $6.00 per 1M output tokens ($3.00 blended at 3:1), list price on 6 Nov 2026. It streams about 71 tokens/s (Artificial Analysis). Released 21 Sep 2026.

Key facts (as of 6 Nov 2026)

Grok 4.7 API pricing: $2.00 per 1M input tokens and $6.00 per 1M output tokens ($3.00 blended at 3:1) (list price, 6 Nov 2026).
Grok 4.7 streams about 71 tokens/s (Artificial Analysis, 6 Nov 2026).
Grok 4.7 was released on 21 Sep 2026 (SpaceXAI).

In detail. Grok 4.7 is a SpaceXAI proprietary reasoning model, released 21 Sep 2026. Its best configuration (xhigh reasoning effort) ranks #125 of 587 on the BenchLeader Index at 58.3 ±2.7, in the upper half. The ± is the point: 217 other configurations score within that range, so they and this one cannot be told apart on quality alone — price and speed are what separate them. It scores highest in long context (63) and lowest in multimodal (56). At $3.00 per million tokens blended it is pricier than most ranked models. Output speed of 71 tokens per second puts it slower than most, with a first answer in 55.1 s. It has been measured at 5 reasoning-effort settings; this summary describes the best-scoring one, and the tabs above switch between them. Last measured 4 Nov 2026.

Blended price: $3.00/M; $2.00 in · $6.00 out; via Artificial Analysis
Output speed: 71 tok/s; measured by Artificial Analysis
First answer: 55 s; first token 1.66 s
Context: 500k; via OpenRouter
Full answer: 147 s; median, reasoning included
Cost per task: $7.77; 24 min a task, AA Index
Tokens per task: 105k; 66% of it reasoning
Released: 21 Sep 2026

How it scores by categoryDashed line = average model (50). One step of 15 = one standard deviation.

Overall index58
Reasoning56
Coding62
Agents & tools57
Maths56
Knowledge61
Human preference62
Multimodal56
Long context63

Versions

SpaceXAI has shipped 11 models under this name. Each is ranked on its own results; a newer version often has fewer results so far, which holds its index nearer the average until more arrive.

Model	Released	Index	Rank
Grok 4.7xhighthis page	21 Sep 2026	58.3	#125
Grok 4.6high	11 Aug 2026	60.5	#74
Grok 4.5	8 Jul 2026	59.8	#90
Grok 4.3high	17 Apr 2026	56.6	#173
Grok 4.2high	10 Mar 2026	–	–
Grok 4.20thinking	17 Feb 2026	58.4	#121
Grok 4.1	17 Nov 2025	57.1	#156
Grok 4	9 Jul 2025	58.3	#124
Grok 3	17 Feb 2025	52.5	#269
Grok 2	12 Dec 2024	42.3	#478
Grok 4.20 1	–	–	–

Reasoning-effort configurations

The same model behaves differently depending on how much it is allowed to think. Each row is one setting, scored only on the benchmarks that were run at that setting. “Not stated” collects results from publishers that did not say which setting they used; for a reasoning model that is usually its thinking mode, but we do not assume it. Pick a setting here or at the top of the page to see its price, speed and category scores.

Effort	Index	Rank	Speed	First answer	Chat reply cost	Categories
low	56.0	#186	71 tok/s	8.87 s	$0.0026	Coding 55 · Knowledge 62 · Long context 65 · Multimodal 61 · Reasoning 47
medium	–	–	65 tok/s	1.66 s	$0.0026	Reasoning 56
high	57.7	#142	71 tok/s	31 s	$0.0026	Coding 61 · Knowledge 63 · Long context 63 · Reasoning 54
xhighbest	58.3	#125	71 tok/s	55 s	$0.0026	Agents & tools 57 · Coding 62 · Human preference 62 · Knowledge 61 · Long context 63 · Maths 56 · Multimodal 56 · Reasoning 56
not stated	–	–	65 tok/s	1.66 s	$0.0026	Agents & tools 54 · Coding 62 · Composite 55 · Maths 43

What more thinking costs

Turning the effort up buys index points and multiplies the bill. Here xhigh costs 2.7× low for +2.3 on the index — 1.6 points per doubling of spend. Compare that with other models.

Each point is one reasoning setting: its cost per Artificial Analysis index task (log scale) and its BenchLeader Index.

How its index has moved

22 Sep 2026 to 6 Nov 2026

xhigh
high
low

The index is recomputed from scratch every day, so a line moves when a new benchmark result lands, when a publisher revises a score, or when the models it is normalised against change. Early movement usually means the score is still settling.

Benchmark results

See all LLM benchmarks. One column per reasoning effort. Rank is among every configuration of every model on that benchmark. Hover a score for the run it came from.

Reasoning

Benchmark	low	medium	high	xhigh	not stated	Source
GPQA Diamond	–	–	–	92.7%#27	–	Epoch AI Benchmarking Hub
Arena Hard Prompts	–	–	–	1463#90	–	Arena (LMArena)
LiveBench 2.0 Reasoningnot in indexeffort not stated	–	–	–	–	82.7%#20	LiveBench
LiveBench 1.0 Reasoningnot in index	–	–	–	82.7%#47	–	LiveBench
Humanity's Last Exam (AA)not in index	39.2%#98	–	42.3%#75	43.1%#65	–	Artificial Analysis
HLE-Diamond (AA)	18.5%#72	–	23.6%#59	23.2%#60	–	Artificial Analysis
ARC-AGI-1	63.3%#159	90.0%#80	90.2%#79	89.5%#83	–	ARC Prize
ARC-AGI-2	16.7%#150	58.8%#93	58.3%#94	61.4%#80	–	ARC Prize
ARC-AGI-3	0.8%#42	8.4%#24	10.0%#22	7.0%#26	–	ARC Prize
CritPtnot in index	13.4%#106	–	18.0%#76	17.7%#78	–	Artificial Analysis
CritPt v2 hard subset (AA)	43.5%#45	–	37.0%#52	33.5%#57	–	Artificial Analysis
MysteryMechanismnot in index	–	–	–	25.2%#17	–	Vals AI
Chess Puzzlesnot in index	–	–	–	41.0%#40	–	Epoch AI Benchmarking Hub
Mystery Game Puzzlesnot in index	–	–	–	29.0%#37	–	Epoch AI Benchmarking Hub
LMCAnot in index	–	–	–	49.4%#69	–	Epoch AI Benchmarking Hub
DTBenchnot in index	–	–	–	96.0%#30	–	Epoch AI Benchmarking Hub

Coding

Benchmark	low	medium	high	xhigh	not stated	Source
SciCode	54.9%#61	–	57.8%#27	57.4%#30	–	SciCode
FrontierCodeeffort not stated	–	–	–	–	47.6%#11	Cognition
Arena Coding	–	–	–	1494#91	–	Arena (LMArena)
Arena WebDev	–	–	–	1638#20	–	Arena (LMArena)
LiveBench 2.0 Codingnot in indexeffort not stated	–	–	–	–	84.5%#17	LiveBench
LiveBench 1.0 Codingnot in index	–	–	–	77.2%#39	–	LiveBench
SciCode (AA)not in index	54.9%#61	–	57.8%#25	57.4%#28	–	Artificial Analysis
IOI	–	–	–	57.7%#19	–	Vals AI
Code Migrationnot in index	–	–	–	44.8%#19	–	Vals AI
Code Migration v1.1not in index	–	–	–	53.2%#12	–	Vals AI
ProgramBenchnot in index	–	–	–	0.5%#18	–	Vals AI
SWE-bench Multimodalnot in index	–	–	–	42.9%#14	–	Vals AI
Vibe Code Bench v1.1not in index	–	–	–	86.2%#13	–	Vals AI
CursorBenchnot in index	33.1%#62	41.6%#38	43.9%#31	46.3%#21	–	Cursor
FrontierSWEnot in index	–	–	–	29.5%#11	–	Epoch AI Benchmarking Hub
Terminal-Bench 4.0 (AA)not in index	16.2%#83	–	24.8%#68	25.8%#65	–	Artificial Analysis
Startrise Frontendnot in indexeffort not stated	–	–	–	–	75.9%#10	Startrise
GameBench 2 Overallnot in indexacross settings	–	–	–	–	63.4#13	DuelLab GameBench 2
GameBench 2 by reasoning settingnot in index	46.7#95	72.0#27	–	71.5#28	–	DuelLab GameBench 2
FrontierCode 1.1 (AA)not in index	39.6%#38	–	34.1%#53	33.8%#54	–	Artificial Analysis
FrontierCode 2.0 (AA)	21.8%#52	–	27.8%#36	28.7%#33	–	Artificial Analysis

Agents & tools

Benchmark	low	high	xhigh	not stated	Source
Terminal-Bench	–	–	37.6%#49	–	Terminal-Bench
APEX-Agentseffort not stated	–	–	–	54.6%#23	Mercor
Arena Agent	–	–	2.7#16	–	Arena (LMArena)
LiveBench Long Agenticnot in indexeffort not stated	–	–	–	46.1%#9	LiveBench
LiveBench Browsernot in indexeffort not stated	–	–	–	49.3%#20	LiveBench
LiveBench Agentic Codingnot in index	–	–	54.0%#34	–	LiveBench
Terminal-Bench 2.1 (Vals)	–	–	73.4%#22	–	Vals AI
GDPval-AA v2.1, normalised scorenot in index	54.1%#28	60.4%#12	60.7%#11	–	Artificial Analysis
GDPval-AA v2.1not in index	1591#30	1716#13	1718#12	–	Artificial Analysis
ITBench SRE (AA)not in index	–	–	42.1%#16	–	Artificial Analysis
Analyst Agent (AA)not in index	–	55.0%#5	52.5%#8	–	Artificial Analysis
CyberBenchnot in index	–	–	69.5%#26	–	Vals AI
Finance Agent v2not in index	–	–	52.3%#37	–	Vals AI
Finance Agent v2.1not in index	–	–	60.6%#21	–	Vals AI
Harvey's Legal Agent Benchmark v1.1not in index	–	–	18.3%#6	–	Vals AI
Harvey's Legal Agent Benchmark v1not in index	–	–	12.5%#14	–	Vals AI
Legal Research Benchnot in index	–	–	47.1%#15	–	Vals AI
Robo-benchnot in index	–	–	0.2%#7	–	Vals AI
SREBenchnot in index	–	–	3.8%#20	–	Vals AI
Tax Agent Benchnot in index	–	–	34.2%#17	–	Vals AI
Terminal-Bench 4.0 (Vals)not in index	–	–	28.8%#25	–	Vals AI
Terminal-Bench Sciencenot in index	–	–	10.0%#16	–	Vals AI
Vending-Bench 2not in indexeffort not stated	–	–	–	10536.8#8	Andon Labs
AutomationBenchnot in index	56.7%#62	63.5%#29	65.6%#22	–	Artificial Analysis
GDP.pdf (AA)not in index	21.4%#66	23.2%#54	20.0%#77	–	Artificial Analysis
Harvey LAB-AA v1.1not in index	–	–	9.4%#3	–	Harvey
AA-Briefcase v1.1not in index	1493#40	1629#14	1637#12	–	Artificial Analysis
GDPxnot in index	–	–	3.9%#15	–	GDPx by SemiAnalysis
Terminal-Bench-Science 0.1 (AA)not in index	4.8%#71	9.5%#56	11.4%#48	–	Artificial Analysis
CWE-Bench (AA)	–	–	68.3%#2	–	Artificial Analysis
DeepSec-Bench (AA)	–	–	26.9%#5	–	Artificial Analysis
CyberGym end-to-end (AA)	–	–	74.0%#8	–	Artificial Analysis

Maths

Benchmark	xhigh	not stated	Source
FrontierMath Tiers 1–3	53.0%#53	–	Epoch AI Benchmarking Hub
FrontierMath Tier 4	17.1%#57	–	Epoch AI Benchmarking Hub
OTIS Mock AIME	98.1%#31	–	Epoch AI Benchmarking Hub
ProofBencheffort not stated	–	26.0%#46	Vals AI
LiveBench Mathematicsnot in index	95.7%#16	–	LiveBench
Arena Maths	1443#93	–	Arena (LMArena)

Knowledge

Benchmark	low	high	xhigh	not stated	Source
SimpleQA Verified	–	–	56.0%#21	–	Epoch AI Benchmarking Hub
LiveBench 2.0 Data Analysisnot in indexeffort not stated	–	–	–	76.9%#14	LiveBench
LiveBench Researchnot in indexeffort not stated	–	–	–	86.4%#2	LiveBench
LiveBench 1.0 Data Analysisnot in index	–	–	76.9%#30	–	LiveBench
AA-Omniscience v1.1	27.4#34	30#29	31.2#26	–	Artificial Analysis
LegalBench	–	–	84.4%#33	–	Vals AI
PRBench Legal	–	–	47.6%#19	–	Scale AI SEAL
BioMysteryBenchnot in index	–	–	69.3%#15	–	Vals AI
Excel Modeling Benchmarknot in index	–	–	67.0%#20	–	Vals AI
Excel Modeling Benchmark v1.1not in index	–	–	71.5%#12	–	Vals AI
MedCodenot in index	–	–	49.5%#22	–	Vals AI
MedScribenot in index	–	–	89.4%#7	–	Vals AI
Public Benefits Bench v1.1not in index	–	–	65.6%#21	–	Vals AI
GDP.pdfnot in index	–	–	22.8%#26	–	Epoch AI Benchmarking Hub
AA-Omniscience v1.1: accuracynot in index	45.7%#42	41.8%#51	42.7%#47	–	Artificial Analysis
AA-Omniscience v1.1: non-hallucinationnot in index	66.2%#33	79.8%#10	80.0%#9	–	Artificial Analysis
HealthBench Professional (AA)not in index	–	–	55.8%#9	–	OpenAI

Instruction following

Benchmark	xhigh	Source
LiveBench Languagenot in index	80.1%#30	LiveBench
LiveBench Instruction Followingnot in index	75.3%#10	LiveBench
Arena Instruction Followingnot in index	1438#84	Arena (LMArena)

Human preference

Benchmark	xhigh	Source
Arena Text	1441#95	Arena (LMArena)
Arena Creative Writingnot in index	1419#86	Arena (LMArena)
Arena Multi-turnnot in index	1432#114	Arena (LMArena)
Arena Longer Queriesnot in index	1452#84	Arena (LMArena)

Multimodal

Benchmark	low	xhigh	not stated	Source
Arena Vision	–	1231#77	–	Arena (LMArena)
LiveBench Designnot in indexeffort not stated	–	–	78.0%#13	LiveBench
MMMU-Pro	78.2%#85	–	–	Artificial Analysis
SAGEnot in index	–	88.7%#7	–	Vals AI
Blueprint-Bench 2not in indexeffort not stated	–	–	32.5%#16	Epoch AI Benchmarking Hub
Code2Video-AAnot in index	–	1127#8	–	Artificial Analysis
AA-3Dnot in index	–	995#13	–	Artificial Analysis

Long context

Benchmark	low	high	xhigh	Source	Trend
AA-LCR v1.1	80.0%#84	77.0%#147	76.7%#149	Artificial Analysis
MLCRnot in index	21.1%#45	18.9%#58	20.0%#50	Artificial Analysis

Composite

Benchmark	low	high	xhigh	not stated	Source
Epoch Capabilities Indexnot in indexeffort not stated	–	–	–	156.1#36	Epoch AI Benchmarking Hub
LiveBench 2.0effort not stated	–	–	–	70.1%#11	LiveBench
LiveBench 1.0not in index	–	–	77.4%#25	–	LiveBench
AA Intelligence Index v5.0not in index	31.1#48	35.2#39	35.5#38	–	Artificial Analysis
AA Intelligence Index v4.3.2not in index	42.2#57	46.3#34	46.5#33	–	Artificial Analysis
AA Progress Index v1not in index	316.4#79	321.6#52	322.7#49	–	Artificial Analysis
Vals Indexnot in index	–	–	58.6#16	–	Vals AI
Vals Finance Indexnot in index	–	–	58.5%#12	–	Vals AI
Vals RSI Indexnot in index	–	–	20.2%#16	–	Vals AI
Vals Science Indexnot in index	–	–	34.8%#15	–	Vals AI
Vals Router Indexnot in indexeffort not stated	–	–	–	58.6%#11	Vals AI
AA Cyber Indexnot in index	–	–	56.4#3	–	Artificial Analysis

GameBench 2 Overall: 63.4. GameBench 2, best reasoning setting (Medium): 72.0. Reference only, not part of the BenchLeader Index.

DuelLab GameBench 2, release gb2-20261103-09b2218cbd85 (https://benchmarks.duellab.org/releases/gb2-20261103-09b2218cbd85/), CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). Scores mapped to BenchLeader model names; not endorsed by DuelLab.

Where it wins

Benchmarks where this configuration ranks in the top five of every configuration measured.

API pricing: where to run it

Every provider serving this model through OpenRouter, with throughput and first-token latency measured on live traffic over the last 30 minutes and each provider’s own price. Purple marks the best in each column.

Provider	Speed	First token	Input $/M	Output $/M	Context	Quantisation
SpaceXAI	78 tok/s	1.63 s	$2.00	$6.00	500k	–
SpaceXAI	74 tok/s	1.60 s	$2.00	$6.00	500k	–
SpaceXAI	62 tok/s	2.01 s	$4.00	$12.00	500k	–
Amazon Bedrock	55 tok/s	3.41 s	$2.00	$6.00	500k	–

What a task costs

Estimates from list price, output speed and time to first answer for the best configuration. “With caching” assumes three-quarters of the input is served from the prompt cache at $0.500 per 1M. Reasoning tokens are not modelled.

Workload	Tokens in / out	Cost	With caching	Time
Chat reply	400 / 300	$0.0026	$0.0022	59.3 s
Summarize a 30-page report	12,000 / 600	$0.028	$0.014	1.1 min
Code edit	6,000 / 1,500	$0.021	$0.014	1.3 min
Agentic coding session	60,000 / 4,000	$0.144	$0.076	1.9 min
Structured extraction	2,000 / 200	$0.0052	$0.0030	57.9 s

New this month

Compare Grok 4.7 with…

Embed or cite

Show where Grok 4.7 ranks on a model card or README. The badge redraws itself with every daily update, carries the data date and links back to this page. To see ranks like this on OpenRouter and Hugging Face model pages, get the Chrome extension.

[![Grok 4.7 on BenchLeader: #62 overall](https://www.benchleader.com/badge/grok-4-7.svg)](https://www.benchleader.com/models/grok-4-7)

Cite this

The ranking changes daily, so a reference carries the data date. APA style:

BenchLeader. (2026, November 6). Grok 4.7 on the BenchLeader Index (method v1.4) [Data set]. Retrieved November 6, 2026, from https://www.benchleader.com/models/grok-4-7

BibTeX:

@misc{benchleader_grok_4_7_20261106,
  author       = {{BenchLeader}},
  title        = {{Grok 4.7 on the BenchLeader Index (method v1.4)}},
  year         = {2026},
  month        = nov,
  howpublished = {\url{https://www.benchleader.com/models/grok-4-7}},
  note         = {Data as of 2026-11-06}
}

Every badge style, the embeddable top-10 widget and how to cite the Index

Questions people ask

What are Grok 4.7's benchmark scores?: As of 6 Nov 2026, Grok 4.7 (xhigh effort) scores 58.3 ±2.7 on the BenchLeader Index, strongest in long context (63) and weakest in multimodal (56). That rests on independent benchmark results only, and every score on this page links to its publisher.
How does Grok 4.7 rank?: Grok 4.7 (xhigh effort) is #125 of 587 ranked configurations on 6 Nov 2026. 217 other configurations score within its margin, so the exact position matters less than the band. The leader is Claude Opus 5.5 (max effort) at 69.3.
How much does Grok 4.7 cost per 1M tokens?: As of 6 Nov 2026, Grok 4.7 costs $2.00 per 1M input tokens and $6.00 per 1M output tokens, $3.00 blended (three input tokens per output token).
Is Grok 4.7 better than DeepSeek V4.1 Flash 1017?: As of 6 Nov 2026, DeepSeek V4.1 Flash 1017 (max) and Grok 4.7 (xhigh) are neck and neck on quality (59.2 vs 58.3 on the BenchLeader Index, inside the ±4.6 margin of error), and DeepSeek V4.1 Flash 1017 (max) costs 5.7× less, which makes it the better buy for most workloads. The head-to-head page has the benchmark-by-benchmark detail.
When was Grok 4.7 released?: SpaceXAI released Grok 4.7 on 21 Sep 2026. BenchLeader has results for it at 5 reasoning-effort settings.

Data as of 6 Nov 2026. Compare these configurations.

Cite as: BenchLeader, “Grok 4.7: benchmarks, pricing, speed and rank”, data as of 6 Nov 2026, https://www.benchleader.com/models/grok-4-7.