Enter opens the highlighted result. Esc closes.

GPT-5 Nano vs Llama 3.1 70B Instruct

GPT-5 Nano leads Llama 3.1 70B Instruct on the Artificial Analysis Intelligence Index, 13.0 to 6.6, and GPT-5 Nano costs 1.4x less per output token ($0.40 vs $0.56 per million). Updated October 22, 2026; 43 benchmarks have a result for both.

Replace GPT-5 Nano Replace Llama 3.1 70B Instruct
GPT-5 NanoOpenAI
Coding
No data
Overall
35.6
Output $/M
$0.40
42
Llama 3.1 70B InstructMeta Llama
Coding
No data
Overall
24.3
Output $/M
$0.56
1
vs boards won of 43 shared

GPT-5 Nano 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-5 Nano for output tokens 1.4x cheaper ($0.40 vs $0.56 per million) and a larger context window (400K against 131K tokens).

Model AGPT-5 NanoOpenAINot benchmarkedNo coding benchmark$0.109 per solved taskModel BLlama 3.1 70B InstructMeta LlamaNot benchmarkedNo coding benchmarkNo cost per solved task
Shared results
43
GPT-5 Nano only
44
Llama 3.1 70B Instruct only
4
Like-for-like categories
7 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-5 Nano not benchmarked, Llama 3.1 70B Instruct not benchmarked.
Widest category
Math, 53.1 points apart on average over 6 shared benchmarks.
Widest single gap
AA Math Index: GPT-5 Nano 83.7 against Llama 3.1 70B Instruct 4.0, 79.7 points apart on a 0 to 100 scale.
Output price
GPT-5 Nano $0.40 against Llama 3.1 70B Instruct $0.56 per million output tokens, 1.4x apart.
Context window
GPT-5 Nano takes 400K 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-5 NanoBarsLlama 3.1 70B InstructTies
Coding50
Intelligence20
Agentic work10
Reasoning60
Math60
Knowledge50
General171
All shared boards421
CodingLike-for-likeSix of these boards feed the Vibe Coding Index50
Coding benchmarks for GPT-5 Nano and Llama 3.1 70B Instruct
BenchmarkGPT-5 NanoLlama 3.1 70B InstructLeader
SWE-bench Verified34.8%Not measuredOnly GPT-5 Nano measured
LiveCodeBench (Vals AI)70.2%Not measuredOnly GPT-5 Nano measured
LiveCodeBench (Artificial Analysis run)78.9%23.2%GPT-5 Nano
Terminal-Bench Hard (Artificial Analysis run)17.4%3.0%GPT-5 Nano
Terminal-Bench21.8%Not measuredOnly GPT-5 Nano measured
WeirdML38.1%9.0%GPT-5 Nano
LMArena Text: coding13511260GPT-5 Nano
LMArena Text: software and IT13451260GPT-5 Nano
Neither model has a result on the other 31 boards here
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index20
Intelligence benchmarks for GPT-5 Nano and Llama 3.1 70B Instruct
BenchmarkGPT-5 NanoLlama 3.1 70B InstructLeader
AA Intelligence Index13.06.6GPT-5 Nano
Epoch Capabilities Index139.2125.9GPT-5 Nano
Agentic workLike-for-likePublic boards, shown as published, kept out of the index10
Agentic work benchmarks for GPT-5 Nano and Llama 3.1 70B Instruct
BenchmarkGPT-5 NanoLlama 3.1 70B InstructLeader
τ²-Bench Telecom36.5%15.2%GPT-5 Nano
BALROGNot measured27.9%Only Llama 3.1 70B Instruct measured
TheAgentCompanyNot measured6.9%Only Llama 3.1 70B Instruct measured
Berkeley Function Calling Leaderboard (BFCL)51.4%Not measuredOnly GPT-5 Nano measured
Neither model has a result on the other 37 boards here
ReasoningLike-for-likePublic boards, shown as published, kept out of the index60
Reasoning benchmarks for GPT-5 Nano and Llama 3.1 70B Instruct
BenchmarkGPT-5 NanoLlama 3.1 70B InstructLeader
GPQA Diamond63.4%Not measuredOnly GPT-5 Nano measured
MMMU Pro (Vals AI)70.9%Not measuredOnly GPT-5 Nano measured
GPQA Diamond (Artificial Analysis run)67.6%40.9%GPT-5 Nano
Humanity's Last Exam (Artificial Analysis run)9.5%4.5%GPT-5 Nano
GPQA Diamond (Epoch AI run)69.4%44.2%GPT-5 Nano
ARC-AGI20.7%Not measuredOnly GPT-5 Nano measured
ARC-AGI-22.6%Not measuredOnly GPT-5 Nano measured
Chess puzzles22.0%Not measuredOnly GPT-5 Nano measured
Mystery game puzzles9.0%Not measuredOnly GPT-5 Nano measured
VPCT37.2%Not measuredOnly GPT-5 Nano measured
LMArena Text: hard prompts13281241GPT-5 Nano
LMArena Text: expert13211209GPT-5 Nano
LMArena Text: hard prompts in English13381261GPT-5 Nano
LMArena Vision: diagrams1141Not measuredOnly GPT-5 Nano measured
LMArena Vision: homework1157Not measuredOnly GPT-5 Nano measured
Kagi LLM Benchmark62.2%Not measuredOnly GPT-5 Nano measured
Neither model has a result on the other 6 boards here
MathLike-for-likePublic boards, shown as published, kept out of the index60
Math benchmarks for GPT-5 Nano and Llama 3.1 70B Instruct
BenchmarkGPT-5 NanoLlama 3.1 70B InstructLeader
AIME81.2%Not measuredOnly GPT-5 Nano measured
MATH 500 (Vals AI)93.8%Not measuredOnly GPT-5 Nano measured
MGSM (Vals AI)89.3%Not measuredOnly GPT-5 Nano measured
AA Math Index83.74.0GPT-5 Nano
AIME 2025 (Artificial Analysis run)83.7%4.0%GPT-5 Nano
OTIS Mock AIME81.1%3.6%GPT-5 Nano
FrontierMath, tiers 1 to 320.0%Not measuredOnly GPT-5 Nano measured
FrontierMath, tier 42.4%Not measuredOnly GPT-5 Nano measured
ProofBench12.0%Not measuredOnly GPT-5 Nano measured
MATH level 595.2%36.7%GPT-5 Nano
LMArena Text: math13171252GPT-5 Nano
LMArena Text: mathematical work13261253GPT-5 Nano
Neither model has a result on the other 1 board here
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index50
Knowledge benchmarks for GPT-5 Nano and Llama 3.1 70B Instruct
BenchmarkGPT-5 NanoLlama 3.1 70B InstructLeader
MMLU Pro76.1%Not measuredOnly GPT-5 Nano measured
LegalBench (Vals AI)50.1%Not measuredOnly GPT-5 Nano measured
CaseLaw (Vals AI)52.6%Not measuredOnly GPT-5 Nano measured
TaxEval (Vals AI)67.4%Not measuredOnly GPT-5 Nano measured
MortgageTax (Vals AI)53.6%Not measuredOnly GPT-5 Nano measured
MedQA (Vals AI)93.3%Not measuredOnly GPT-5 Nano measured
MedCode (Vals AI)30.4%Not measuredOnly GPT-5 Nano measured
MMLU Pro (Artificial Analysis run)78.0%67.6%GPT-5 Nano
SimpleQA Verified11.7%Not measuredOnly GPT-5 Nano measured
MMLUNot measured80.1%Only Llama 3.1 70B Instruct measured
LMArena Text: science13181276GPT-5 Nano
LMArena Text: medicine and healthcare13351257GPT-5 Nano
LMArena Text: legal and government12981283GPT-5 Nano
LMArena Text: business and finance13111235GPT-5 Nano
Neither model has a result on the other 4 boards here
Web designDirectional onlyPublic boards, shown as published, kept out of the index6 boards, none shared
Web design benchmarks for GPT-5 Nano and Llama 3.1 70B Instruct
BenchmarkGPT-5 NanoLlama 3.1 70B InstructLeader
Design Arena: websites1101Not measuredOnly GPT-5 Nano measured
Design Arena: UI components1064Not measuredOnly GPT-5 Nano measured
Design Arena: game dev1044Not measuredOnly GPT-5 Nano measured
Design Arena: data viz1056Not measuredOnly GPT-5 Nano measured
Design Arena: 3D965Not measuredOnly GPT-5 Nano measured
Design Arena: all code categories1084Not measuredOnly GPT-5 Nano measured
Neither model has a result on the other 3 boards here
GeneralLike-for-likePublic boards, shown as published, kept out of the index171
General benchmarks for GPT-5 Nano and Llama 3.1 70B Instruct
BenchmarkGPT-5 NanoLlama 3.1 70B InstructLeader
SAGE (Vals AI)30.4%Not measuredOnly GPT-5 Nano measured
MedScribe (Vals AI)72.9%Not measuredOnly GPT-5 Nano measured
IFBench67.6%34.4%GPT-5 Nano
Long context reasoning (AA-LCR)45.0%Not measuredOnly GPT-5 Nano measured
DTBench62.7%60.0%GPT-5 Nano
LMCA7.9%14.8%Llama 3.1 70B Instruct
Fiction.LiveBench44.4%Not measuredOnly GPT-5 Nano measured
LMArena Text13201261GPT-5 Nano
LMArena Text: instruction following13061231GPT-5 Nano
LMArena Text: creative writing12491232GPT-5 Nano
LMArena Vision1159Not measuredOnly GPT-5 Nano measured
LMArena Text: English13211294GPT-5 Nano
LMArena Text: non-English13131219GPT-5 Nano
LMArena Text: Chinese13561215GPT-5 Nano
LMArena Text: German13271222GPT-5 Nano
LMArena Text: Japanese12271132GPT-5 Nano
LMArena Text: Korean12691140GPT-5 Nano
LMArena Text: FrenchNot measured1262Only Llama 3.1 70B Instruct measured
LMArena Text: Spanish13601253GPT-5 Nano
LMArena Text: Russian12961234GPT-5 Nano
LMArena Text: Polish1314Not measuredOnly GPT-5 Nano measured
LMArena Text: multi-turn13111256GPT-5 Nano
LMArena Text: longer queries13121241GPT-5 Nano
LMArena Text: writing and language12831236GPT-5 Nano
LMArena Text: entertainment and media12681230GPT-5 Nano
LMArena Vision: English1170Not measuredOnly GPT-5 Nano measured
LMArena Vision: Chinese1176Not measuredOnly GPT-5 Nano measured
LMArena Vision: reading text1152Not measuredOnly GPT-5 Nano measured
LMArena Vision: humor1110Not measuredOnly GPT-5 Nano measured
Neither model has a result on the other 17 boards here
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Image generation benchmarks for GPT-5 Nano and Llama 3.1 70B Instruct
BenchmarkGPT-5 NanoLlama 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-5 Nano$0.0003Fits in one request
Llama 3.1 70B Instruct$0.0008Fits in one request

GPT-5 Nano costs less

Repository review

50K input + 3K output tokens

GPT-5 Nano$0.0037Fits in one request
Llama 3.1 70B Instruct$0.030Fits in one request

GPT-5 Nano costs less

Agent loop

200K input + 10K output tokens

GPT-5 Nano$0.014Fits in one request
Llama 3.1 70B Instruct$0.118Exceeds the context window

GPT-5 Nano 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-5 Nano and Llama 3.1 70B Instruct
SpecificationOpenAI GPT-5 NanoMeta Llama 3.1 70B Instruct
MakerOpenAIMeta Llama
WeightsNo public weights listedOpen weights
ReleasedAugust 7, 2025July 23, 2024
Context window400K131K
Input $/Mtok$0.05$0.56
Output $/Mtok$0.40$0.56
Coding indexNot benchmarkedNot benchmarked
Independent measurements behind it00
Cost per solved task$0.109Not computable
SourceOpenRouter listingOpenRouter listing

Questions

Is GPT-5 Nano 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-5 Nano for output tokens 1.4x cheaper ($0.40 vs $0.56 per million) and a larger context window (400K against 131K tokens).

Which is cheaper, GPT-5 Nano or Llama 3.1 70B Instruct?

GPT-5 Nano costs $0.05 per million input tokens and $0.40 per million output tokens; Llama 3.1 70B Instruct costs $0.56 and $0.56. Cost per solved task is known for GPT-5 Nano only.

Which has the bigger context window, GPT-5 Nano or Llama 3.1 70B Instruct?

GPT-5 Nano, with 400K tokens against 131K.

How many benchmarks is the GPT-5 Nano vs Llama 3.1 70B Instruct comparison based on?

43 benchmarks have a result for both models as of October 22, 2026. GPT-5 Nano'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