Grok 4.3

+ contribute

Grok 4.3 is a closed-weights language model from SpaceXAI (xAI), released April 2026.

70.2±0.9#80 of 719 LLMsfrom 15 benchmark results across 6 sources

$ how it compares

The LLMs ranked either side of Grok 4.3, out of 828 with a score, then more from SpaceXAI (xAI).

#	model	score	where	RAM	released
83	Ornith 1.5 35B-A3B DeepReinforce	71±2	local	~22GB	2026-08
84	MiniMax-M3 MiniMax	70±1	api $0.48/M +local	~137GB	2026-06
85	Grok 4.3 SpaceXAI (xAI)	70±1	api $1.56/M		2026-04
86	Gemini 2.5 Pro Preview (06-05) Google	70±2	comparison only		2025-06
87	o3-pro OpenAI	70±2	api $35.00/M		2025-06
	more from SpaceXAI (xAI)
30	Grok 4.6 SpaceXAI (xAI)	77±1	api $3.00/M		2026-08
42	Grok 4.5 SpaceXAI (xAI)	75±1	api $3.00/M		2026-07
47	Grok 4.7 SpaceXAI (xAI)	75±1	api $3.00/M		2026-09

see all LLMs · find the ones that fit your machine

$ what it's good at

agent tasks▁▁▃▆█▆▅▅▅▄▃▁ #108 of 300 · 11% of its score · 4 tests

Trails SmolLM3 3B, Nova 2 Pro (Preview), Mistral Medium 3.5 (128B) here, though they rank lower overall.

test	result	vs the others in its run	difficulty	counts	why
Agents' Last Exam overall full-pass rate (fixed OpenClaw + GUI-as-Tool)	4.3% ✎	▄···█▄▄▄█▄·█ beats 0/10	very hard · avg 1.7%	▇▇ 4.6%	typical
Thinkingbox-bench pass@1 (507 stateful business-workflow agent tasks, N=20)	14.4% ✎	▅··▃·▃·█▃▅██ beats 2/15	very hard · avg 4.9%	▇ 2.8%	at its level
Legal Research Bench (agentic, all-pass)	15.3% ✎	▄█··▄▄▄██·▄█ beats 1/12	very hard · avg 5.8%	▇ 2.1%	typical
KlinikeBench strict pass@1 (333 interactive clinical encounters, Terminus-2)	21.3% ✎	▁▁·▁▁▁▁▅▃█▅█ beats 19/27	very hard · avg 2.4%	▇ 1.4%	typical

coding▁▁▁▃▆██▇▆▄▃▁ #111 of 458 * · 12% of its score · 1 test

Trails Llama 3.1 Nemotron Ultra 253B v1, Solar Pro 4, Laguna XS 2.1 (33B-A3B) here, though they rank lower overall.

test	result	vs the others in its run	difficulty	counts	why
Arena WebDev (Code Arena)	1356.6 ✎	▁▂▄▃▅█▄▇▆▄▁▁ beats 30/119	avg 981	▇▇▇▇▇ 12%	sharp

chat & writing▁▁▁▃▅▄▅█▆▆▇▃ #61 of 393 * · 6.6% of its score · 1 test

Beats GPT-5.2, Qwen3.8 27B, LongCat 2.0 (1.6T-A48B) here, though they rank higher overall. Trails Gemma 4 31B, Gemini 2.5 Pro, Claude Sonnet 4.5 here, though they rank lower overall.

test	result	vs the others in its run	difficulty	counts	why
LMArena (text)	1441.4 ✎	▁▁▁▂▃▃▃▆▆▇█▃ beats 300/379	avg 1268	▇▇▇ 6.6%	blunt

math▁▂▃▅▆▅▄█▄▁▁▁ #60 of 398 · 47% of its score · 3 tests

Beats Qwen3.6 Max Preview, Claude Sonnet 4.6, Claude Opus 4.5 here, though they rank higher overall. Trails INTELLECT-3, GLM-4.7, A.X K2 (688B-A33B) here, though they rank lower overall.

test	result	vs the others in its run	difficulty	counts	why
FrontierMath tiers 1–3 v2	42.8% ✎	▂··▁▁▃▄▄█▅▃▃ beats 29/82	very hard · avg 1.2%	▇▇▇▇▇▇▇▇▇▇▇▇▇ 31%	sharp · at its level
FrontierMath tier 4 v2	14.6% ✎	▁▂▁▂█▅▂▂▁▁▁▁ beats 15/63	very hard · avg 0.1%	▇▇▇▇▇ 12%	sharp
OTIS mock AIME 2024–2025	93.3% ✎	▃▃▄▃▃▃▇▇█▃▂▅ beats 153/193	very hard · avg 8.5%	▇ 3.0%	sharp · noisy

vision & documents▁▁▃▂▁▃▄▂▆▇█▄ #52 of 147 * · 4.5% of its score · 1 test

Beats DeepSeek V4 Flash Vision (Exp), GPT-5, Claude Opus 4.5 here, though they rank higher overall. Trails Qwen3.6 Flash, Qwen3.5 35B-A3B, Gemma 4 31B here, though they rank lower overall.

test	result	vs the others in its run	difficulty	counts	why
Arena vision (style controlled)	1242.9 ✎	▁▁▂▂▂▂▄▅▅▆██ beats 97/145	avg 1068	▇▇ 4.5%	blunt

web search▂▁▂▆▄█▃▃▁▁·▁ #63 of 80 * · 4.8% of its score · 2 tests

Beats DeepSeek V4 Flash (Preview, Apr 2026) here, though they rank higher overall. Trails GPT-4o mini (2024-07-18), Mistral Small 3.1 24B (2503), Llama 4 Scout (109B-A17B) here, though they rank lower overall.

test	result	vs the others in its run	difficulty	counts	why
Arena search systems (style controlled)	1193.2 ✎	▁·▁▁▂▂▂█▅▂▁▁ beats 15/30	avg 1055	▇ 2.7%	blunt
Legal Research Bench (agentic, all-pass)	15.3% ✎	▄█··▄▄▄██·▄█ beats 1/12	very hard · avg 5.8%	▇ 2.1%	typical

other tests4 not in a task · 16% of its score

test	result	vs the others in its run	difficulty	counts	why
GPQA Diamond	88.8% ✎	▁▁▁▁▂▄▆▄▅▇█▅ beats 166/208	hard · avg 40%	▇▇▇ 7.9%	sharp
Chess Puzzles (Epoch)	25.0% ✎	█▁▁▁▁▂▁▄▂▂▁▁ beats 95/135	very hard · avg 5.9%	▇ 3.7%	sharp · noisy
SimpleQA Verified	33.2% ✎	▁▁▃▂▃▃▆▆█▃▂▅ beats 24/78	very hard · avg 8.0%	▇ 2.5%	at its level
Legal Research Bench (single-shot, no tools, all-pass)	4.1% ✎	▃▃█▃·▅···█▃▃ beats 6/12	very hard · avg 0.8%	▇ 2.3%	too hard for it

$ where to run it

Through an API

api · xAI · $1.56/M · $1.25 in / $2.5 out per 1M tokens · 1M context · docs.x.ai ↗ · price sourceStandard base token rate from models.dev. Provider-specific tiers, long-context surcharges, caching and promotions may differ.

$ how the score was worked out · Epoch AI — internal evaluatio… 61%, LMArena leaderboard dataset 26%, Agents' Last Exam — authors'… 5% and 3 more

Each result implies a score on its own, after adjusting for how hard its test is and who ran it. The score is their weighted average, with a small pull toward the middle for thinly tested models. Each test's weight is its counts column above; here it's added up by who ran the tests. How it works

source	results	share of the score	licence
Epoch AI — internal evaluations	6	▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 61%	CC BY 4.0
LMArena leaderboard dataset	4	▇▇▇▇▇▇▇▇ 26%	CC BY 4.0
Agents' Last Exam — authors' own fixed-harness model comparison	1	▇ 5%	CC BY 4.0
Legal Research Bench (Vals AI, arXiv:2610.00609v1)	2	▇ 4%	CC-BY-4.0
One Success Isn't Reliability: Thinkingbox, a Sandbox and Benchmark for Agents in Stateful Business Workflows (arXiv:2608.19741v4)	1	▇ 3%	CC BY 4.0
KlinikeBench: Evaluating Language Models Beyond Diagnostic Accuracy (arXiv:2609.38480v2)	1	▇ 1%	CC BY 4.0
prior · pull toward the middle		▇ 0%
weighted average		= 70.2