Noometry

Model comparison

Claude Sonnet 4.6 vs Llama-3.3-70B-Instruct

Claude Sonnet 4.6 is the stronger model overall, scoring 50.1 to 30.3 on the Noometry Index. Llama-3.3-70B-Instruct costs 39× less per token, which makes it the better buy when Claude Sonnet 4.6's lead doesn't matter for your workload.

Last verified . 27 shared benchmarks.

Claude Sonnet 4.6 Anthropic

50.1

Rank #42 Confirmed

Llama-3.3-70B-Instruct Meta

30.3

Rank #283 Confirmed

Summary

  • They share 27 benchmarks with published results for both. Claude Sonnet 4.6 scores higher in 9 categories and Llama-3.3-70B-Instruct in 0 categories; 9 gaps are clear of the uncertainty.
  • The widest gap is in math, where Claude Sonnet 4.6 leads 51.9 to 14.3.
  • The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 85.8% for Claude Sonnet 4.6 and 5.1% for Llama-3.3-70B-Instruct.
  • Llama-3.3-70B-Instruct is cheaper at $0.10 / $0.32 per million input/output tokens, against $3 / $15 for Claude Sonnet 4.6.
  • Claude Sonnet 4.6 accepts more context: 1M tokens versus 128K.
  • Llama-3.3-70B-Instruct has downloadable open weights; the other is API-only.

Side by side

Claude Sonnet 4.6Llama-3.3-70B-Instruct
ProviderAnthropicMeta
Noometry Index50.130.3
Released2026-02-172024-12-06
WeightsProprietaryOpen
Context window1M128K
Max output128K4K
Input $ / M tokens$3$0.10
Output $ / M tokens$15$0.32
Results tracked5743

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

Category by category

Coding Claude Sonnet 4.6 leads

Claude Sonnet 4.6: 46.0 (#62), Llama-3.3-70B-Instruct: 30.6 (#284)

BenchmarkClaude Sonnet 4.6Llama-3.3-70B-Instruct
SciCode46.8%26%
WeirdML66.1%14.4%
LMArena Coding15041268
SWE-bench Verified75.2%—
DeepSWE29.9%—
FrontierCode24.3%—
LMArena WebDev1521—
BigCodeBench Instruct—46.9%
LiveBench Coding—36.6%
BigCodeBench Complete—57.5%
ALE-Bench1,327—

Agentic & Tool Use Claude Sonnet 4.6 leads

Claude Sonnet 4.6: 39.1 (#26), Llama-3.3-70B-Instruct: 25.6 (#100)

BenchmarkClaude Sonnet 4.6Llama-3.3-70B-Instruct
Terminal-Bench53.4%—
APEX-Agents43%—
Berkeley Function Calling Leaderboard—31.9%
OSWorld 2.09.3%—
DeepResearch Bench54.9%—
OSWorld72.1%—
BALROG—23%
ExploitBench23.6%—
GBAEval48.8%—
GDP.pdf18%—
LMArena Search1221—
Vending-Bench 27,204—

Reasoning Claude Sonnet 4.6 leads

Claude Sonnet 4.6: 46.1 (#40), Llama-3.3-70B-Instruct: 13.7 (#325)

BenchmarkClaude Sonnet 4.6Llama-3.3-70B-Instruct
CritPt3.1%0%
LMArena Hard Prompts14851257
DTBench89.9%59.5%
LMCA46.5%17.5%
Epoch Capabilities Index152.25127.32
ForecastBench6258.6
ARC-AGI-260.4%—
SimpleBench—19.9%
NYT Connections (extended)80.9%—
ARC-AGI-186.5%—
Chess Puzzles13%—
Thematic Generalization76.3%—
LiveBench Reasoning—50.8%
Mystery Game Puzzles16%—
LiveBench Data Analysis—49.5%
LiveBench—50.2%

Math Claude Sonnet 4.6 leads

Claude Sonnet 4.6: 51.9 (#41), Llama-3.3-70B-Instruct: 14.3 (#292)

BenchmarkClaude Sonnet 4.6Llama-3.3-70B-Instruct
OTIS Mock AIME 2024-202585.8%5.1%
LMArena Math14631268
ProofBench45%—
LiveBench Math—42.2%
MATH Level 5—41.6%
FrontierMath (Feb 2025 set)32.4%—
FrontierMath Tier 4 (v1)8.3%—

Knowledge Claude Sonnet 4.6 leads

Claude Sonnet 4.6: 51.6 (#60), Llama-3.3-70B-Instruct: 30.5 (#220)

BenchmarkClaude Sonnet 4.6Llama-3.3-70B-Instruct
GPQA Diamond87.4%47.4%
Vectara Hallucination Rate10.6%4.1%
LMArena Expert15001225
SimpleQA Verified35.5%—
Confabulations—22.8%
MMLU—86.3%

Multimodal Not comparable

Claude Sonnet 4.6: 38.0 (#61), Llama-3.3-70B-Instruct: —

BenchmarkClaude Sonnet 4.6Llama-3.3-70B-Instruct
LMArena Vision1283—
Blueprint-Bench 26.7%—
LMArena Document1482—

Multilingual Claude Sonnet 4.6 leads

Claude Sonnet 4.6: 54.6 (#39), Llama-3.3-70B-Instruct: 40.1 (#215)

BenchmarkClaude Sonnet 4.6Llama-3.3-70B-Instruct
LMArena Non-English14401236
LMArena Chinese14911217
LMArena French14651280
LMArena German14281251
LMArena Japanese14191149
LMArena Korean14111142
LMArena Russian14401252
LMArena Spanish14651269

Instruction Following Claude Sonnet 4.6 leads

Claude Sonnet 4.6: 77.6 (#23), Llama-3.3-70B-Instruct: 71.3 (#152)

BenchmarkClaude Sonnet 4.6Llama-3.3-70B-Instruct
LMArena Instruction Following14761242
LiveBench Instruction Following—82.7%

Long Context Claude Sonnet 4.6 leads

Claude Sonnet 4.6: 45.3 (#40), Llama-3.3-70B-Instruct: 26.4 (#290)

BenchmarkClaude Sonnet 4.6Llama-3.3-70B-Instruct
LMArena Longer Query14791256
Fiction.LiveBench—33.3%

Writing & Preference Claude Sonnet 4.6 leads

Claude Sonnet 4.6: 70.4 (#22), Llama-3.3-70B-Instruct: 47.8 (#203)

BenchmarkClaude Sonnet 4.6Llama-3.3-70B-Instruct
LMArena Text14581274
LMArena Creative Writing14351250
LMArena Multi-Turn14641279
EQ-Bench Creative Writing1810—
EQ-Bench 41207—
LiveBench Language—39.2%

Frequently asked questions

Is Claude Sonnet 4.6 better than Llama-3.3-70B-Instruct?

Claude Sonnet 4.6 is the stronger model overall, scoring 50.1 to 30.3 on the Noometry Index. Llama-3.3-70B-Instruct costs 39× less per token, which makes it the better buy when Claude Sonnet 4.6's lead doesn't matter for your workload.

Which is cheaper, Claude Sonnet 4.6 or Llama-3.3-70B-Instruct?

Llama-3.3-70B-Instruct is cheaper. It lists at $0.10 per million input tokens and $0.32 per million output tokens; Claude Sonnet 4.6 lists at $3 and $15.

Is Claude Sonnet 4.6 or Llama-3.3-70B-Instruct better for coding?

Claude Sonnet 4.6 scores higher on coding benchmarks: 46.0 versus 30.6 in the Noometry coding category.

Which has the bigger context window?

Claude Sonnet 4.6 does, with 1M tokens against 128K.

How many benchmarks do Claude Sonnet 4.6 and Llama-3.3-70B-Instruct share?

27 benchmarks have published results for both models. Claude Sonnet 4.6 has 57 scored results on Noometry and Llama-3.3-70B-Instruct has 43.

Related comparisons

Go deeper