Noometry

Model comparison

Claude Opus 4.8 vs Grok 4.20 (Non-Reasoning)

Claude Opus 4.8 is the stronger model overall, scoring 60.7 to 48.5 on the Noometry Index. Grok 4.20 (Non-Reasoning) costs 6.4× less per token, which makes it the better buy when Claude Opus 4.8's lead doesn't matter for your workload.

Last verified . 42 shared benchmarks.

Claude Opus 4.8 Anthropic

60.7

Rank #9 Confirmed

Grok 4.20 (Non-Reasoning) xAI

48.5

Rank #48 Confirmed

Summary

  • They share 42 benchmarks with published results for both. Claude Opus 4.8 scores higher in 9 categories and Grok 4.20 (Non-Reasoning) in 1 category; 8 gaps are clear of the uncertainty.
  • The widest gap is in math, where Claude Opus 4.8 leads 78.2 to 48.0.
  • The biggest single-benchmark swing is ProofBench: 69% for Claude Opus 4.8 and 14% for Grok 4.20 (Non-Reasoning).
  • Grok 4.20 (Non-Reasoning) is cheaper at $1.25 / $2.50 per million input/output tokens, against $5 / $25 for Claude Opus 4.8.

Side by side

Claude Opus 4.8Grok 4.20 (Non-Reasoning)
ProviderAnthropicxAI
Noometry Index60.748.5
Released2026-05-282026-02-17
WeightsProprietaryProprietary
Context window1M1M
Max output128K30K
Input $ / M tokens$5$1.25
Output $ / M tokens$25$2.50
Results tracked6546

Sponsored placements are available on pages like this one. Advertise on Noometry

Category by category

Coding Claude Opus 4.8 leads

Claude Opus 4.8: 60.2 (#9), Grok 4.20 (Non-Reasoning): 41.8 (#104)

BenchmarkClaude Opus 4.8Grok 4.20 (Non-Reasoning)
LMArena WebDev15561374
WeirdML82.9%52.3%
LMArena Coding14911458
ALE-Bench1,5641,150
DeepSWE59%—
FrontierCode46.5%—
SciCode53.5%—
GSO47.1%—

Agentic & Tool Use Claude Opus 4.8 leads

Claude Opus 4.8: 47.6 (#10), Grok 4.20 (Non-Reasoning): 34.3 (#43)

BenchmarkClaude Opus 4.8Grok 4.20 (Non-Reasoning)
τ²-bench Banking39.7%18%
LMArena Search12041189
Vending-Bench 25,7874,663
Terminal-Bench—57.3%
APEX-Agents48.9%—
OSWorld 2.020.6%—
Remote Labor Index8.3%—
DeepResearch Bench50.2%—
PostTrainBench33.8%—
GBAEval70.9%—
GDP.pdf24%—

Reasoning Claude Opus 4.8 leads

Claude Opus 4.8: 64.6 (#14), Grok 4.20 (Non-Reasoning): 52.2 (#30)

BenchmarkClaude Opus 4.8Grok 4.20 (Non-Reasoning)
ARC-AGI-272.1%65.1%
Kagi LLM Benchmark88.8%75%
NYT Connections (extended)91.1%85.4%
ARC-AGI-192.5%89.5%
Chess Puzzles34%24%
LMArena Hard Prompts14811451
DTBench94.9%90.1%
LMCA57.5%38.7%
Epoch Capabilities Index158.3152.01
ForecastBench59.961.4
SimpleBench64.8%—
CritPt20.9%—
EnigmaEval23.5%—
Thematic Generalization—63.8%
EBR-Bench28.6%—
Mystery Game Puzzles36%—
Surface Evolver Bench87.5%—
Bench to the Future 30.14—

Math Claude Opus 4.8 leads

Claude Opus 4.8: 78.2 (#11), Grok 4.20 (Non-Reasoning): 48.0 (#54)

BenchmarkClaude Opus 4.8Grok 4.20 (Non-Reasoning)
FrontierMath (Tiers 1-3)80%44.9%
FrontierMath Tier 456.1%17.1%
OTIS Mock AIME 2024-202598.3%92.2%
ProofBench69%14%
LMArena Math14871455
MathArena Final-Answer Competitions91.8%—
FrontierMath (Feb 2025 set)47.2%—
FrontierMath Tier 4 (v1)31.3%—

Knowledge Claude Opus 4.8 leads

Claude Opus 4.8: 61.3 (#26), Grok 4.20 (Non-Reasoning): 52.8 (#54)

BenchmarkClaude Opus 4.8Grok 4.20 (Non-Reasoning)
GPQA Diamond91%89.3%
SimpleQA Verified53%30.2%
LMArena Expert15011440

Multimodal Claude Opus 4.8 leads

Claude Opus 4.8: 43.2 (#20), Grok 4.20 (Non-Reasoning): 33.2 (#88)

BenchmarkClaude Opus 4.8Grok 4.20 (Non-Reasoning)
LMArena Vision12951263
Blueprint-Bench 214.5%0%
LMArena Document14751416
Furniture Assembly42.5%—

Multilingual Too close to call

Claude Opus 4.8: 55.4 (#31), Grok 4.20 (Non-Reasoning): 54.7 (#38)

BenchmarkClaude Opus 4.8Grok 4.20 (Non-Reasoning)
LMArena Non-English14501440
LMArena Chinese15091481
LMArena French14821477
LMArena German14701463
LMArena Japanese14401450
LMArena Korean14321417
LMArena Russian14761457
LMArena Spanish14701442

Instruction Following Claude Opus 4.8 leads

Claude Opus 4.8: 77.6 (#22), Grok 4.20 (Non-Reasoning): 75.0 (#79)

BenchmarkClaude Opus 4.8Grok 4.20 (Non-Reasoning)
LMArena Instruction Following14761420

Long Context Too close to call

Claude Opus 4.8: 45.4 (#37), Grok 4.20 (Non-Reasoning): 45.4 (#33)

BenchmarkClaude Opus 4.8Grok 4.20 (Non-Reasoning)
LMArena Longer Query14821436
CL-bench—22.2%
CL-bench Life—11.9%

Writing & Preference Claude Opus 4.8 leads

Claude Opus 4.8: 72.1 (#17), Grok 4.20 (Non-Reasoning): 65.9 (#42)

BenchmarkClaude Opus 4.8Grok 4.20 (Non-Reasoning)
LMArena Text14601451
LMArena Creative Writing14531439
EQ-Bench Creative Writing18401574
LMArena Multi-Turn14761456
EQ-Bench 41281—

Frequently asked questions

Is Claude Opus 4.8 better than Grok 4.20 (Non-Reasoning)?

Claude Opus 4.8 is the stronger model overall, scoring 60.7 to 48.5 on the Noometry Index. Grok 4.20 (Non-Reasoning) costs 6.4× less per token, which makes it the better buy when Claude Opus 4.8's lead doesn't matter for your workload.

Which is cheaper, Claude Opus 4.8 or Grok 4.20 (Non-Reasoning)?

Grok 4.20 (Non-Reasoning) is cheaper. It lists at $1.25 per million input tokens and $2.50 per million output tokens; Claude Opus 4.8 lists at $5 and $25.

Is Claude Opus 4.8 or Grok 4.20 (Non-Reasoning) better for coding?

Claude Opus 4.8 scores higher on coding benchmarks: 60.2 versus 41.8 in the Noometry coding category.

Which has the bigger context window?

Both accept 1M tokens.

How many benchmarks do Claude Opus 4.8 and Grok 4.20 (Non-Reasoning) share?

42 benchmarks have published results for both models. Claude Opus 4.8 has 65 scored results on Noometry and Grok 4.20 (Non-Reasoning) has 46.

Related comparisons

Go deeper