When was Grok 4.6 released?

Grok 4.6 was released by SpaceXAI on Wednesday, Aug 12 2026.

Grok 4.6 was built by SpaceXAI. Elon Musk's AI company building the Grok series of models. Founded in 2023 as xAI, now part of SpaceX. Cursor (Anysphere) and its Composer coding models were acquired in 2026 and are tracked here.

How much does Grok 4.6 cost?

Grok 4.6 costs $2.00 per million input tokens and $6.00 per million output tokens through the SpaceXAI API. Cached input is $0.50 per million tokens. Those are the rates for the “Under 200K input tokens” tier; 1 other pricing tier is published for this model. Rates are pay-as-you-go API prices verified against SpaceXAI's published pricing on August 18, 2026.

What benchmark scores did Grok 4.6 get?

Grok 4.6 reports 13 tracked benchmark scores — BullshitBench v2: 66%; DeepSWE 1.1: 65.9%; FrontierCode v1.1 (Extended) (extended split): 61.3%; APEX-SWE: 56.4%; Next.js Evals: 71%; Terminal-Bench 4.0: 20.3%; Terminal-Bench 3.0: 26%; APEX-Agents: 57.5%; Harvey's Legal Agent Benchmark: 15.8%; AA Intelligence Index: 61; GDPval-AA v2: 1753; AA-Briefcase: 1577; threejseval: 1519. Scores are the figures published at release by SpaceXAI. It holds the best score among all models tracked here on APEX-SWE and APEX-Agents.

Is Grok 4.6 open source?

No. Grok 4.6 is a proprietary model. The weights are not published — it is available only through the provider's own API, apps, or partner platforms.

What came before and after Grok 4.6?

SpaceXAI's previous tracked release was Grok 4.5 on Jul 8 2026, 35 days earlier. It is the most recent SpaceXAI model tracked on AI Release Tracker.

Grok 4.6

Released Aug 12 2026

Grok 4.6 is an AI model released by SpaceXAI on Wednesday, Aug 12 2026, 35 days after Grok 4.5. Benchmark results (shown below) cover APEX-SWE, APEX-Agents, BullshitBench v2, DeepSWE 1.1, FrontierCode v1.1 (Extended), Next.js Evals, and 7 more.

API pricing

Under 200K input tokens

Input: $2.00
Cached input: $0.50
Output: $6.00

At least 200K input tokens

Input: $4.00
Cached input: $1.00
Output: $12.00

Reasoning or thinking is supported.

USD per 1M tokens

Available from


Amazon Bedrock	us-west-2	$2.20	$6.60	500K

USD per 1M tokens

Benchmarks

Coding

APEX-SWE

56.4%

DeepSWE 1.1

65.9%

#11 of 26

FrontierCode v1.1 (Extended)

61.3%

extended split

#2 of 3

Next.js Evals

71%

#11 of 27

Terminal & CLI

Terminal-Bench 4.0

20.3%

via Terminal-Bench

#11 of 16

Terminal-Bench 3.0

26%

#4 of 7

Agentic & tool use

APEX-Agents

57.5%

Knowledge work

AA Intelligence Index

#2 of 6

GDPval-AA v2

1753

#7 of 20

AA-Briefcase

1577

#3 of 4

Legal & tax

Harvey's Legal Agent Benchmark

15.8%

#2 of 8

Robustness

BullshitBench v2

66%

#18 of 77

Community preference

threejseval

1519

#8 of 16

Source: Terminal-Bench, retrieved 30 August 2026. Every other score here is the figure the lab published at launch.

About

Grok 4.6, released August 12, 2026, was xAI's bet that post-training alone could deliver a flagship upgrade. It reused the 1.5-trillion-parameter base model of Grok 4.5, five weeks its senior, and poured the gains into supervised fine-tuning on regenerated trajectories and wide-ranging reinforcement learning across engineering and domain-specific environments. At release it scored 61 on the Artificial Analysis Intelligence Index — up from 56 for Grok 4.5, and level with GPT-5.6 Sol at the time — while priced at $2 per million input tokens and $6 per million output.

xAI pitched it at long-running agents and knowledge work rather than raw coding leaderboards. At launch it posted a GDPval-AA v2 Elo of 1753, 15.8% on Harvey's Legal Agent Benchmark — the top score in xAI's own launch comparison. Terminal work stayed a relative weakness, at 26% on the much harder Terminal-Bench 3.0. Elon Musk said the larger Grok 4.7 would follow within weeks, trading some serving speed for capability.

Compare Grok 4.6 with

Grok 4.6

Suggested comparisons

Grok 4.6vsGrok 4.5 Grok 4.6vsGPT-6 Astra Grok 4.6vsClaude Fable 5.1 Grok 4.6vsGemini 3.8 Flash Grok 4.6vsMuse Spark 1.3 Grok 4.6vsDeepSeek-V4.1-Flash Grok 4.6vsMistral Medium 3.5 Grok 4.6vsKimi K3 Grok 4.6vsGLM-5.3-Flash Grok 4.6vsQwen3.8-Max-0902 Grok 4.6vsNemotron 3.5 Lightning

Frequently asked questions

: Grok 4.6 was released by SpaceXAI on Wednesday, Aug 12 2026.
: Grok 4.6 was built by SpaceXAI. Elon Musk's AI company building the Grok series of models. Founded in 2023 as xAI, now part of SpaceX. Cursor (Anysphere) and its Composer coding models were acquired in 2026 and are tracked here.
: Grok 4.6 costs $2.00 per million input tokens and $6.00 per million output tokens through the SpaceXAI API. Cached input is $0.50 per million tokens. Those are the rates for the “Under 200K input tokens” tier; 1 other pricing tier is published for this model. Rates are pay-as-you-go API prices verified against SpaceXAI's published pricing on August 18, 2026.
: Grok 4.6 reports 13 tracked benchmark scores — BullshitBench v2: 66%; DeepSWE 1.1: 65.9%; FrontierCode v1.1 (Extended) (extended split): 61.3%; APEX-SWE: 56.4%; Next.js Evals: 71%; Terminal-Bench 4.0: 20.3%; Terminal-Bench 3.0: 26%; APEX-Agents: 57.5%; Harvey's Legal Agent Benchmark: 15.8%; AA Intelligence Index: 61; GDPval-AA v2: 1753; AA-Briefcase: 1577; threejseval: 1519. Scores are the figures published at release by SpaceXAI. It holds the best score among all models tracked here on APEX-SWE and APEX-Agents.
: No. Grok 4.6 is a proprietary model. The weights are not published — it is available only through the provider's own API, apps, or partner platforms.
: SpaceXAI's previous tracked release was Grok 4.5 on Jul 8 2026, 35 days earlier. It is the most recent SpaceXAI model tracked on AI Release Tracker.