Enter opens the highlighted result. Esc closes.

GPT-4 vs Llama 3.1 70B Instruct

GPT-4 leads Llama 3.1 70B Instruct on the Artificial Analysis Intelligence Index, 6.7 to 6.6, and Llama 3.1 70B Instruct costs 107.1x less per output token ($0.56 vs $60.00 per million). Updated October 21, 2026; 35 benchmarks have a result for both.

Replace GPT-4 Replace Llama 3.1 70B Instruct
GPT-4OpenAI
Coding
No data
Overall
6.7
Output $/M
$60.00
5
Llama 3.1 70B InstructMeta Llama
Coding
No data
Overall
6.6
Output $/M
$0.56
30
vs boards won of 35 shared

GPT-4 or Llama 3.1 70B Instruct: which to pick

Neither model has a coding score yet, so the two cannot be ranked on coding.

Pick GPT-4 for work measured by LMArena Text: coding or LMArena Text: software and IT, where it leads.

Model AGPT-4OpenAINot benchmarkedNo coding benchmarkNo cost per solved taskModel BLlama 3.1 70B InstructMeta LlamaNot benchmarkedNo coding benchmarkNo cost per solved task
Shared results
35
GPT-4 only
10
Llama 3.1 70B Instruct only
12
Like-for-like categories
6 of 9

Only benchmarks that tested both models count toward a pick; a result only one model has is shown but not compared.

Which one for your work

Each card is decided by one number both models carry. Where only one has it, the card says so and makes no call.

At a glance

Coding index
GPT-4 not benchmarked, Llama 3.1 70B Instruct not benchmarked.
Widest category
Knowledge, 13.9 points apart on average over 6 shared benchmarks.
Widest single gap
LMArena Text: medicine and healthcare: GPT-4 1145 against Llama 3.1 70B Instruct 1257, 18.7 points apart on a 0 to 100 scale.
Output price
GPT-4 $60.00 against Llama 3.1 70B Instruct $0.56 per million output tokens, 107.1x apart.
Context window
GPT-4 takes 8K tokens, Llama 3.1 70B Instruct 131K.

Benchmark by benchmark

First the score: boards each model leads, per category, where both have a result. Then every board, a category to a fold; bars are each result on a 0 to 100 scale, so compare the two in a row, not one row with another.

Boards won, where both have a result
CategoryGPT-4BarsLlama 3.1 70B InstructTies
Coding02
Intelligence11
Reasoning04
Math02
Knowledge15
General316
All shared boards530
CodingLike-for-likeSix of these boards feed the Vibe Coding Index02
Coding benchmarks for GPT-4 and Llama 3.1 70B Instruct
BenchmarkGPT-4Llama 3.1 70B InstructLeader
SWE-bench Verified22.4%Not measuredOnly GPT-4 measured
SWE-bench Lite39.7%Not measuredOnly GPT-4 measured
SWE-bench (full test set)12.5%Not measuredOnly GPT-4 measured
AA Coding Index13.1Not measuredOnly GPT-4 measured
LiveCodeBench (Artificial Analysis run)Not measured23.2%Only Llama 3.1 70B Instruct measured
Terminal-Bench Hard (Artificial Analysis run)Not measured3.0%Only Llama 3.1 70B Instruct measured
WeirdMLNot measured9.0%Only Llama 3.1 70B Instruct measured
LMArena Text: coding12091260Llama 3.1 70B Instruct
LMArena Text: software and IT11951260Llama 3.1 70B Instruct
Neither model has a result on the other 29 boards here
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index11
Intelligence benchmarks for GPT-4 and Llama 3.1 70B Instruct
BenchmarkGPT-4Llama 3.1 70B InstructLeader
AA Intelligence Index6.76.6GPT-4
Epoch Capabilities Index125.7125.9Llama 3.1 70B Instruct
Agentic workDirectional onlyPublic boards, shown as published, kept out of the index3 boards, none shared
Agentic work benchmarks for GPT-4 and Llama 3.1 70B Instruct
BenchmarkGPT-4Llama 3.1 70B InstructLeader
τ²-Bench TelecomNot measured15.2%Only Llama 3.1 70B Instruct measured
BALROGNot measured27.9%Only Llama 3.1 70B Instruct measured
TheAgentCompanyNot measured6.9%Only Llama 3.1 70B Instruct measured
Neither model has a result on the other 36 boards here
ReasoningLike-for-likePublic boards, shown as published, kept out of the index04
Reasoning benchmarks for GPT-4 and Llama 3.1 70B Instruct
BenchmarkGPT-4Llama 3.1 70B InstructLeader
GPQA Diamond (Artificial Analysis run)34.9%40.9%Llama 3.1 70B Instruct
Humanity's Last Exam (Artificial Analysis run)Not measured4.5%Only Llama 3.1 70B Instruct measured
GPQA Diamond (Epoch AI run)Not measured44.2%Only Llama 3.1 70B Instruct measured
BIG-Bench Hard75.1%Not measuredOnly GPT-4 measured
LMArena Text: hard prompts12001241Llama 3.1 70B Instruct
LMArena Text: expert11491209Llama 3.1 70B Instruct
LMArena Text: hard prompts in English12091261Llama 3.1 70B Instruct
Neither model has a result on the other 14 boards here
MathLike-for-likePublic boards, shown as published, kept out of the index02
Math benchmarks for GPT-4 and Llama 3.1 70B Instruct
BenchmarkGPT-4Llama 3.1 70B InstructLeader
AA Math IndexNot measured4.0Only Llama 3.1 70B Instruct measured
AIME 2025 (Artificial Analysis run)Not measured4.0%Only Llama 3.1 70B Instruct measured
OTIS Mock AIMENot measured3.6%Only Llama 3.1 70B Instruct measured
MATH level 5Not measured36.7%Only Llama 3.1 70B Instruct measured
GSM8K92.0%Not measuredOnly GPT-4 measured
LMArena Text: math12301252Llama 3.1 70B Instruct
LMArena Text: mathematical work12281253Llama 3.1 70B Instruct
Neither model has a result on the other 6 boards here
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index15
Knowledge benchmarks for GPT-4 and Llama 3.1 70B Instruct
BenchmarkGPT-4Llama 3.1 70B InstructLeader
MMLU Pro (Artificial Analysis run)56.2%67.6%Llama 3.1 70B Instruct
MMLU86.4%80.1%GPT-4
AI2 Reasoning Challenge96.3%Not measuredOnly GPT-4 measured
LMArena Text: science11771276Llama 3.1 70B Instruct
LMArena Text: medicine and healthcare11451257Llama 3.1 70B Instruct
LMArena Text: legal and government11911283Llama 3.1 70B Instruct
LMArena Text: business and finance11451235Llama 3.1 70B Instruct
Neither model has a result on the other 10 boards here
Web designNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Web design benchmarks for GPT-4 and Llama 3.1 70B Instruct
BenchmarkGPT-4Llama 3.1 70B InstructLeader
Neither model has a result on 9 boards here
GeneralLike-for-likePublic boards, shown as published, kept out of the index316
General benchmarks for GPT-4 and Llama 3.1 70B Instruct
BenchmarkGPT-4Llama 3.1 70B InstructLeader
IFBench33.2%34.4%Llama 3.1 70B Instruct
DTBench62.7%60.0%GPT-4
LMCA17.1%14.8%GPT-4
HellaSwag95.3%Not measuredOnly GPT-4 measured
WinoGrande87.5%Not measuredOnly GPT-4 measured
LMArena Text12061261Llama 3.1 70B Instruct
LMArena Text: instruction following12011231Llama 3.1 70B Instruct
LMArena Text: creative writing11911232Llama 3.1 70B Instruct
LMArena Text: English12201294Llama 3.1 70B Instruct
LMArena Text: non-English11941219Llama 3.1 70B Instruct
LMArena Text: Chinese11841215Llama 3.1 70B Instruct
LMArena Text: German11981222Llama 3.1 70B Instruct
LMArena Text: Japanese11371132GPT-4
LMArena Text: Korean10871140Llama 3.1 70B Instruct
LMArena Text: French12201262Llama 3.1 70B Instruct
LMArena Text: Spanish11971253Llama 3.1 70B Instruct
LMArena Text: Russian11951234Llama 3.1 70B Instruct
LMArena Text: Polish1265Not measuredOnly GPT-4 measured
LMArena Text: multi-turn12061256Llama 3.1 70B Instruct
LMArena Text: longer queries11901241Llama 3.1 70B Instruct
LMArena Text: writing and language12191236Llama 3.1 70B Instruct
LMArena Text: entertainment and media11841230Llama 3.1 70B Instruct
Neither model has a result on the other 24 boards here
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Image generation benchmarks for GPT-4 and Llama 3.1 70B Instruct
BenchmarkGPT-4Llama 3.1 70B InstructLeader
Neither model has a result on 5 boards here

What each workload costs

Three fixed token mixes at each model's listed price, so per-token rates become a bill.

Chat turn

1K input + 500 output tokens

GPT-4$0.060Fits in one request
Llama 3.1 70B Instruct$0.0008Fits in one request

Llama 3.1 70B Instruct costs less

Repository review

50K input + 3K output tokens

GPT-4$1.68Exceeds the context window
Llama 3.1 70B Instruct$0.030Fits in one request

Llama 3.1 70B Instruct costs less

Agent loop

200K input + 10K output tokens

GPT-4$6.60Exceeds the context window
Llama 3.1 70B Instruct$0.118Exceeds the context window

Llama 3.1 70B Instruct costs less

Listed OpenRouter rates, without cache or batch discounts. A workload that does not fit a model's context window would need splitting.

Specification differences

The leading value in a row is marked; a row with a missing value names no leader. Our two rows are marked in red.

Specifications of GPT-4 and Llama 3.1 70B Instruct
SpecificationOpenAI GPT-4Meta Llama 3.1 70B Instruct
MakerOpenAIMeta Llama
WeightsNo public weights listedOpen weights
ReleasedMay 28, 2023July 23, 2024
Context window8K131K
Input $/Mtok$30.00$0.56
Output $/Mtok$60.00$0.56
Coding indexNot benchmarkedNot benchmarked
Independent measurements behind it00
Cost per solved taskNot computableNot computable
SourceOpenRouter listingOpenRouter listing

Questions

Is GPT-4 better than Llama 3.1 70B Instruct for coding?

Neither model has a coding score yet, so the two cannot be ranked on coding. Pick GPT-4 for work measured by LMArena Text: coding or LMArena Text: software and IT, where it leads.

Which is cheaper, GPT-4 or Llama 3.1 70B Instruct?

GPT-4 costs $30.00 per million input tokens and $60.00 per million output tokens; Llama 3.1 70B Instruct costs $0.56 and $0.56. Cost per solved task is known for neither of them.

Which has the bigger context window, GPT-4 or Llama 3.1 70B Instruct?

Llama 3.1 70B Instruct, with 131K tokens against 8K.

How many benchmarks is the GPT-4 vs Llama 3.1 70B Instruct comparison based on?

35 benchmarks have a result for both models as of October 21, 2026. GPT-4's coding index rests on 0 independent benchmarks and Llama 3.1 70B Instruct's on 0. Where only one model has a result, the page says so instead of naming a winner.

Compare next