gpt-oss-20b vs Llama 3.1 70B Instruct
gpt-oss-20b leads Llama 3.1 70B Instruct on the Artificial Analysis Intelligence Index, 10.0 to 6.6, and gpt-oss-20b costs 3.7x less per output token ($0.15 vs $0.56 per million). Updated October 22, 2026; 42 benchmarks have a result for both.
gpt-oss-20b 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-oss-20b for output tokens 3.7x cheaper ($0.15 vs $0.56 per million) and work measured by LMArena Text: legal and government or LMCA, where it leads.
- Shared results
- 42
- gpt-oss-20b only
- 20
- Llama 3.1 70B Instruct only
- 5
- 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.
Coding
Vibe Coding Index, six public coding benchmarks
No call
Neither has a coding score yet.
One side not measured
Intelligence
AA Intelligence Index
gpt-oss-20b
gpt-oss-20b leads, 10.0 against 6.6.
Same board, both measured
Reasoning
GPQA Diamond
No call
Only gpt-oss-20b has a GPQA Diamond result, 68.9%.
Directional only
Math
AIME
No call
Only gpt-oss-20b has a AIME result, 86.0%.
Directional only
Knowledge
MMLU Pro
No call
Only gpt-oss-20b has a MMLU Pro result, 71.6%.
Directional only
Web design
Design Arena: websites
No call
Only gpt-oss-20b has a Design Arena: websites result, 852.
Directional only
Long documents
Listed context window
No call
Both take 131K tokens.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
gpt-oss-20b
gpt-oss-20b $0.0001, Llama 3.1 70B Instruct $0.0008 per turn.
List prices
At a glance
- Coding index
- gpt-oss-20b not benchmarked, Llama 3.1 70B Instruct not benchmarked.
- Widest category
- Math, 50.4 points apart on average over 5 shared benchmarks.
- Widest single gap
- AIME 2025 (Artificial Analysis run): gpt-oss-20b 89.3% against Llama 3.1 70B Instruct 4.0%, 85.3 points apart on a 0 to 100 scale.
- Output price
- gpt-oss-20b $0.15 against Llama 3.1 70B Instruct $0.56 per million output tokens, 3.7x apart.
- Context window
- Both take 131K tokens.
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.
| Category | gpt-oss-20b | Bars | Llama 3.1 70B Instruct | Ties |
|---|---|---|---|---|
| Coding | 5 | 0 | ||
| Intelligence | 2 | 0 | ||
| Agentic work | 1 | 0 | ||
| Reasoning | 6 | 0 | ||
| Math | 5 | 0 | ||
| Knowledge | 4 | 1 | ||
| General | 15 | 3 | ||
| All shared boards | 38 | 4 |
CodingLike-for-likeSix of these boards feed the Vibe Coding Index50
| Benchmark | gpt-oss-20b | Llama 3.1 70B Instruct | Leader |
|---|---|---|---|
| LiveCodeBench (Vals AI) | 80.4% | Not measured | Only gpt-oss-20b measured |
| LiveCodeBench (Artificial Analysis run) | 77.7% | 23.2% | gpt-oss-20b |
| SciCode | 38.9% | Not measured | Only gpt-oss-20b measured |
| Terminal-Bench Hard (Artificial Analysis run) | 10.6% | 3.0% | gpt-oss-20b |
| Terminal-Bench | 3.4% | Not measured | Only gpt-oss-20b measured |
| WeirdML | 40.9% | 9.0% | gpt-oss-20b |
| ALE-Bench | 566.0 | Not measured | Only gpt-oss-20b measured |
| LMArena Text: coding | 1307 | 1260 | gpt-oss-20b |
| LMArena Text: software and IT | 1311 | 1260 | gpt-oss-20b |
| Neither model has a result on the other 30 boards here | |||
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index20
| Benchmark | gpt-oss-20b | Llama 3.1 70B Instruct | Leader |
|---|---|---|---|
| AA Intelligence Index | 10.0 | 6.6 | gpt-oss-20b |
| Epoch Capabilities Index | 137.7 | 125.9 | gpt-oss-20b |
Agentic workLike-for-likePublic boards, shown as published, kept out of the index10
| Benchmark | gpt-oss-20b | Llama 3.1 70B Instruct | Leader |
|---|---|---|---|
| τ²-Bench Telecom | 60.2% | 15.2% | gpt-oss-20b |
| BALROG | Not measured | 27.9% | Only Llama 3.1 70B Instruct measured |
| TheAgentCompany | Not measured | 6.9% | Only Llama 3.1 70B Instruct measured |
| BrowseComp-Plus | 50.6% | Not measured | Only gpt-oss-20b measured |
| Neither model has a result on the other 37 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index60
| Benchmark | gpt-oss-20b | Llama 3.1 70B Instruct | Leader |
|---|---|---|---|
| GPQA Diamond | 68.9% | Not measured | Only gpt-oss-20b measured |
| GPQA Diamond (Artificial Analysis run) | 68.8% | 40.9% | gpt-oss-20b |
| Humanity's Last Exam (Artificial Analysis run) | 11.0% | 4.5% | gpt-oss-20b |
| GPQA Diamond (Epoch AI run) | 60.8% | 44.2% | gpt-oss-20b |
| LMArena Text: hard prompts | 1274 | 1241 | gpt-oss-20b |
| LMArena Text: expert | 1258 | 1209 | gpt-oss-20b |
| LMArena Text: hard prompts in English | 1292 | 1261 | gpt-oss-20b |
| Kagi LLM Benchmark | 53.2% | Not measured | Only gpt-oss-20b measured |
| Neither model has a result on the other 14 boards here | |||
MathLike-for-likePublic boards, shown as published, kept out of the index50
| Benchmark | gpt-oss-20b | Llama 3.1 70B Instruct | Leader |
|---|---|---|---|
| AIME | 86.0% | Not measured | Only gpt-oss-20b measured |
| MATH 500 (Vals AI) | 94.2% | Not measured | Only gpt-oss-20b measured |
| MGSM (Vals AI) | 89.0% | Not measured | Only gpt-oss-20b measured |
| AA Math Index | 89.3 | 4.0 | gpt-oss-20b |
| AIME 2025 (Artificial Analysis run) | 89.3% | 4.0% | gpt-oss-20b |
| OTIS Mock AIME | 65.3% | 3.6% | gpt-oss-20b |
| MATH level 5 | Not measured | 36.7% | Only Llama 3.1 70B Instruct measured |
| LMArena Text: math | 1317 | 1252 | gpt-oss-20b |
| LMArena Text: mathematical work | 1305 | 1253 | gpt-oss-20b |
| Neither model has a result on the other 4 boards here | |||
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index41
| Benchmark | gpt-oss-20b | Llama 3.1 70B Instruct | Leader |
|---|---|---|---|
| MMLU Pro | 71.6% | Not measured | Only gpt-oss-20b measured |
| LegalBench (Vals AI) | 70.8% | Not measured | Only gpt-oss-20b measured |
| CaseLaw (Vals AI) | 43.8% | Not measured | Only gpt-oss-20b measured |
| CorpFin (Vals AI) | 53.1% | Not measured | Only gpt-oss-20b measured |
| TaxEval (Vals AI) | 63.7% | Not measured | Only gpt-oss-20b measured |
| MedQA (Vals AI) | 82.9% | Not measured | Only gpt-oss-20b measured |
| MMLU Pro (Artificial Analysis run) | 74.8% | 67.6% | gpt-oss-20b |
| MMLU | Not measured | 80.1% | Only Llama 3.1 70B Instruct measured |
| LMArena Text: science | 1296 | 1276 | gpt-oss-20b |
| LMArena Text: medicine and healthcare | 1310 | 1257 | gpt-oss-20b |
| LMArena Text: legal and government | 1280 | 1283 | Llama 3.1 70B Instruct |
| LMArena Text: business and finance | 1284 | 1235 | gpt-oss-20b |
| Neither model has a result on the other 6 boards here | |||
Web designDirectional onlyPublic boards, shown as published, kept out of the index2 boards, none shared
| Benchmark | gpt-oss-20b | Llama 3.1 70B Instruct | Leader |
|---|---|---|---|
| Design Arena: websites | 852 | Not measured | Only gpt-oss-20b measured |
| Design Arena: data viz | 933 | Not measured | Only gpt-oss-20b measured |
| Neither model has a result on the other 7 boards here | |||
GeneralLike-for-likePublic boards, shown as published, kept out of the index153
| Benchmark | gpt-oss-20b | Llama 3.1 70B Instruct | Leader |
|---|---|---|---|
| IFBench | 65.1% | 34.4% | gpt-oss-20b |
| Long context reasoning (AA-LCR) | 34.7% | Not measured | Only gpt-oss-20b measured |
| DTBench | 68.0% | 60.0% | gpt-oss-20b |
| LMCA | 14.5% | 14.8% | Llama 3.1 70B Instruct |
| LMArena Text | 1288 | 1261 | gpt-oss-20b |
| LMArena Text: instruction following | 1236 | 1231 | gpt-oss-20b |
| LMArena Text: creative writing | 1201 | 1232 | Llama 3.1 70B Instruct |
| LMArena Text: English | 1303 | 1294 | gpt-oss-20b |
| LMArena Text: non-English | 1269 | 1219 | gpt-oss-20b |
| LMArena Text: Chinese | 1315 | 1215 | gpt-oss-20b |
| LMArena Text: German | 1255 | 1222 | gpt-oss-20b |
| LMArena Text: Japanese | 1244 | 1132 | gpt-oss-20b |
| LMArena Text: Korean | 1236 | 1140 | gpt-oss-20b |
| LMArena Text: French | Not measured | 1262 | Only Llama 3.1 70B Instruct measured |
| LMArena Text: Spanish | 1267 | 1253 | gpt-oss-20b |
| LMArena Text: Russian | 1278 | 1234 | gpt-oss-20b |
| LMArena Text: Polish | 1252 | Not measured | Only gpt-oss-20b measured |
| LMArena Text: multi-turn | 1268 | 1256 | gpt-oss-20b |
| LMArena Text: longer queries | 1250 | 1241 | gpt-oss-20b |
| LMArena Text: writing and language | 1238 | 1236 | gpt-oss-20b |
| LMArena Text: entertainment and media | 1216 | 1230 | Llama 3.1 70B Instruct |
| Neither model has a result on the other 25 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | gpt-oss-20b | Llama 3.1 70B Instruct | Leader |
|---|---|---|---|
| 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-oss-20b costs less
Repository review
50K input + 3K output tokens
gpt-oss-20b costs less
Agent loop
200K input + 10K output tokens
gpt-oss-20b 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.
| Specification | OpenAI gpt-oss-20b | Meta Llama 3.1 70B Instruct |
|---|---|---|
| Maker | OpenAI | Meta Llama |
| Weights | Open weights | Open weights |
| Released | August 5, 2025 | July 23, 2024 |
| Context window | 131K | 131K |
| Input $/Mtok | $0.03 | $0.56 |
| Output $/Mtok | $0.15 | $0.56 |
| Coding index | Not benchmarked | Not benchmarked |
| Independent measurements behind it | 0 | 0 |
| Cost per solved task | Not computable | Not computable |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is gpt-oss-20b 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-oss-20b for output tokens 3.7x cheaper ($0.15 vs $0.56 per million) and work measured by LMArena Text: legal and government or LMCA, where it leads.
Which is cheaper, gpt-oss-20b or Llama 3.1 70B Instruct?
gpt-oss-20b costs $0.03 per million input tokens and $0.15 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-oss-20b or Llama 3.1 70B Instruct?
Neither: both take 131K tokens.
How many benchmarks is the gpt-oss-20b vs Llama 3.1 70B Instruct comparison based on?
42 benchmarks have a result for both models as of October 22, 2026. gpt-oss-20b'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.