Grok 4.6 vs Grok4.5
Wins 5 of 8 areas
Coding · Agents · Reasoning · Math · Long documents
Wins 1 of 8 areas
Images and charts
Grok 4.6 is the stronger all-rounder.Grok4.5 is better at images and charts.
Scores updated · 33 tests both models report · How we compare
Where each one wins
Tests won in each of the eight areas we test. Each piece is one test, so a longer bar means more evidence; grey means the two scored within a point of each other.
- ReasoningHard problems that need careful thinking60Grok 4.66 of 7 tests · 1 tie
- AgentsCarrying out multi-step tasks on its own30Grok 4.63 of 3 tests
- MathCompetition and research-level math30Grok 4.63 of 3 tests
- CodingWriting and fixing software31Grok 4.63 of 5 tests · 1 tie
- Long documentsFinding answers in very long texts10Grok 4.61 of 1 test
- Images and chartsUnderstanding pictures, charts and video01Grok4.51 of 1 test
- FactsGetting facts right instead of making them up11Even1 each · 1 tie
- Following instructionsDoing exactly what it is asked00Even0 each · 1 tie
Images and charts, long documents and following instructions rest on a single test each.
What it costs
Prices per million tokens, roughly 750,000 words. The bars show the cost of a million tokens read plus a million written.
The biggest differences
The tests each model wins by the widest margin, up to three each. Scores are out of 100.
Where Grok 4.6 pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+30.1points ahead
- Abstract visual puzzles that people can solveARC-AGI-2+14.5points ahead
- Real work tasks from 44 professionsGDPVal+12.8points ahead
Where Grok4.5 pulls ahead
- Answers hard knowledge questions correctlyAA-Omniscience · Accuracy+8.6points ahead
- Code for real scientific research problemsSciCode+2points ahead
Every test, side by side
All 33 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingGrok 4.6
- LiveBench · CodingGrok 4.6 by 8.276.868.6+8.2
- LMArena · WebDevGrok 4.6 by 64 rating points16171553+64 rating
- Terminal-Bench 2.1Grok 4.6 by 6.48881.6+6.4
- SciCodeGrok4.5 by 25355+2
- LiveBench · Agentic Codingtie5756.5tie
AgentsGrok 4.6
- GDPValGrok 4.6 by 12.857.344.5+12.8
- Terminal-Bench 4.0Grok 4.6 by 6.617.210.6+6.6
- τ-Bench V3 · BankingGrok 4.6 by 1.243.342.1+1.2
ReasoningGrok 4.6
- ARC-AGI-2Grok 4.6 by 14.567.152.6+14.5
- SimpleBenchGrok 4.6 by 5.975.970+5.9
- CritPtGrok 4.6 by 4.319.715.4+4.3
- LiveBench · ReasoningGrok 4.6 by 3.390.587.2+3.3
- ARC-AGI-3Grok 4.6 by 1.82.10.3+1.8
- Humanity's Last ExamGrok 4.6 by 1.444.142.7+1.4
- GPQA Diamondtie93.593.1tie
FactsEven
- AA-Omniscience · Non-hallucinationGrok 4.6 by 30.17645.9+30.1
- AA-Omniscience · AccuracyGrok4.5 by 8.64351.5+8.6
- SimpleQA Verifiedtie48.948.3tie
Images and chartsGrok4.5
- LMArena · VisionGrok4.5 by 23 rating points12651288+23 rating
MathGrok 4.6
- FrontierMath Tiers 1-3 (v2)Grok 4.6 by 8.86657.2+8.8
- FrontierMath Tier 4Grok 4.6 by 7.331.724.4+7.3
- LiveBench · MathematicsGrok 4.6 by 1.892.690.8+1.8
Following instructionsEven
- LiveBench · Instruction Followingtie71.971.5tie
Other results9 tests, not counted
Tests outside the eight areas. They are not counted above: several are summary scores built from other tests, or the same test under another name.
- DeepSWE 1.1Grok 4.6 by 11.965.954+11.9
- AA Agentic IndexGrok 4.6 by 10.652.742.1+10.6
- Terminal-Bench 4.0Grok 4.6 by 7.920.312.4+7.9
- AA IntelligenceGrok 4.6 by 5.444.238.8+5.4
- AA-OmniscienceGrok 4.6 by 429.325.3+4
- Artificial Analysis Coding IndexGrok 4.6 by 3.575.972.4+3.5
- ARC-AGI-1Grok 4.6 by 1.38785.7+1.3
- livebench_data_analysistie73.973tie
- livebench_languagetie83.782.8tie
Questions people ask
Which is better, Grok 4.6 or Grok4.5?
Grok 4.6 wins five of the eight areas we test: coding, agents, reasoning, math and long documents. Grok4.5 wins images and charts. They are level on facts and following instructions.
Which is better for coding?
Grok 4.6. It wins 3 of the 5 coding tests both models report; Grok4.5 wins 1, and 1 is a tie.
Which is cheaper?
Grok 4.6 costs $2.00 per million input tokens and $6.00 per million output tokens; Grok4.5 costs $2.00 and $6.00. For a million tokens read plus a million written, they cost about the same.
How do you compare the two?
We use the 33 benchmark tests both models have published scores on. The verdict counts the 24 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 9 are listed but not counted, because several are summary scores or repeat a test. Each score is the one shown on the model's own page.