GPT-4o-mini (2024-07-18) vs Llama 3.3 70B Instruct
GPT-4o-mini (2024-07-18) costs 1.2x less per output token than Llama 3.3 70B Instruct ($0.60 vs $0.72 per million). Updated October 22, 2026; 35 benchmarks have a result for both.
GPT-4o-mini (2024-07-18) or Llama 3.3 70B Instruct: which to pick
Neither model has a coding score yet, so the two cannot be ranked on coding.
Pick GPT-4o-mini (2024-07-18) for output tokens 1.2x cheaper ($0.60 vs $0.72 per million) and work measured by WeirdML or BALROG, where it leads.
- Shared results
- 35
- GPT-4o-mini (2024-07-18) only
- 11
- Llama 3.3 70B Instruct only
- 17
- 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.
Coding
Vibe Coding Index, six public coding benchmarks
No call
Neither has a coding score yet.
One side not measured
Intelligence
AA Intelligence Index
No call
Only Llama 3.3 70B Instruct has a AA Intelligence Index result, 7.7.
Directional only
Long documents
Listed context window
Llama 3.3 70B Instruct
Llama 3.3 70B Instruct takes 131K tokens against 128K.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
GPT-4o-mini (2024-07-18)
GPT-4o-mini (2024-07-18) $0.0004, Llama 3.3 70B Instruct $0.0008 per turn.
List prices
At a glance
- Coding index
- GPT-4o-mini (2024-07-18) not benchmarked, Llama 3.3 70B Instruct not benchmarked.
- Widest category
- Agentic work, 5.6 points apart on average over 1 shared benchmark.
- Widest single gap
- MATH level 5: GPT-4o-mini (2024-07-18) 52.6% against Llama 3.3 70B Instruct 41.6%, 11.0 points apart on a 0 to 100 scale.
- Output price
- GPT-4o-mini (2024-07-18) $0.60 against Llama 3.3 70B Instruct $0.72 per million output tokens, 1.2x apart.
- Context window
- GPT-4o-mini (2024-07-18) takes 128K tokens, Llama 3.3 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.
| Category | GPT-4o-mini (2024-07-18) | Bars | Llama 3.3 70B Instruct | Ties |
|---|---|---|---|---|
| Coding | 2 | 1 | ||
| Agentic work | 0 | 1 | ||
| Reasoning | 3 | 2 | ||
| Math | 4 | 0 | ||
| Knowledge | 4 | 1 | ||
| General | 15 | 2 | ||
| All shared boards | 28 | 7 |
CodingLike-for-likeSix of these boards feed the Vibe Coding Index21
| Benchmark | GPT-4o-mini (2024-07-18) | Llama 3.3 70B Instruct | Leader |
|---|---|---|---|
| Aider polyglot | 3.6% | Not measured | Only GPT-4o-mini (2024-07-18) measured |
| AA Coding Index | Not measured | 11.9 | Only Llama 3.3 70B Instruct measured |
| LiveCodeBench (Artificial Analysis run) | Not measured | 28.8% | Only Llama 3.3 70B Instruct measured |
| Terminal-Bench Hard (Artificial Analysis run) | Not measured | 3.0% | Only Llama 3.3 70B Instruct measured |
| WeirdML | 11.8% | 14.4% | Llama 3.3 70B Instruct |
| LMArena Text: coding | 1290 | 1268 | GPT-4o-mini (2024-07-18) |
| LMArena Text: software and IT | 1291 | 1273 | GPT-4o-mini (2024-07-18) |
| Neither model has a result on the other 36 boards here | |||
IntelligenceDirectional onlyPublic boards, shown as published, kept out of the index2 boards, none shared
| Benchmark | GPT-4o-mini (2024-07-18) | Llama 3.3 70B Instruct | Leader |
|---|---|---|---|
| AA Intelligence Index | Not measured | 7.7 | Only Llama 3.3 70B Instruct measured |
| Epoch Capabilities Index | Not measured | 127.3 | Only Llama 3.3 70B Instruct measured |
Agentic workLike-for-likePublic boards, shown as published, kept out of the index01
| Benchmark | GPT-4o-mini (2024-07-18) | Llama 3.3 70B Instruct | Leader |
|---|---|---|---|
| τ²-Bench Telecom | Not measured | 26.6% | Only Llama 3.3 70B Instruct measured |
| BALROG | 17.4% | 23.0% | Llama 3.3 70B Instruct |
| Berkeley Function Calling Leaderboard (BFCL) | Not measured | 31.9% | Only Llama 3.3 70B Instruct measured |
| Neither model has a result on the other 39 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index32
| Benchmark | GPT-4o-mini (2024-07-18) | Llama 3.3 70B Instruct | Leader |
|---|---|---|---|
| GPQA Diamond (Artificial Analysis run) | Not measured | 49.8% | Only Llama 3.3 70B Instruct measured |
| Humanity's Last Exam (Artificial Analysis run) | Not measured | 3.6% | Only Llama 3.3 70B Instruct measured |
| GPQA Diamond (Epoch AI run) | 37.7% | 47.4% | Llama 3.3 70B Instruct |
| SimpleBench | 10.7% | 19.9% | Llama 3.3 70B Instruct |
| Chess puzzles | 3.0% | Not measured | Only GPT-4o-mini (2024-07-18) measured |
| Mystery game puzzles | 12.0% | Not measured | Only GPT-4o-mini (2024-07-18) measured |
| VPCT | 34.0% | Not measured | Only GPT-4o-mini (2024-07-18) measured |
| LMArena Text: hard prompts | 1267 | 1257 | GPT-4o-mini (2024-07-18) |
| LMArena Text: expert | 1234 | 1225 | GPT-4o-mini (2024-07-18) |
| LMArena Text: hard prompts in English | 1278 | 1277 | GPT-4o-mini (2024-07-18) |
| Neither model has a result on the other 13 boards here | |||
MathLike-for-likePublic boards, shown as published, kept out of the index40
| Benchmark | GPT-4o-mini (2024-07-18) | Llama 3.3 70B Instruct | Leader |
|---|---|---|---|
| AA Math Index | Not measured | 7.7 | Only Llama 3.3 70B Instruct measured |
| AIME 2025 (Artificial Analysis run) | Not measured | 7.7% | Only Llama 3.3 70B Instruct measured |
| OTIS Mock AIME | 6.9% | 5.1% | GPT-4o-mini (2024-07-18) |
| FrontierMath, tiers 1 to 3 | 0.7% | Not measured | Only GPT-4o-mini (2024-07-18) measured |
| MATH level 5 | 52.6% | 41.6% | GPT-4o-mini (2024-07-18) |
| LMArena Text: math | 1267 | 1267 | GPT-4o-mini (2024-07-18) |
| LMArena Text: mathematical work | 1271 | 1267 | GPT-4o-mini (2024-07-18) |
| Neither model has a result on the other 7 boards here | |||
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index41
| Benchmark | GPT-4o-mini (2024-07-18) | Llama 3.3 70B Instruct | Leader |
|---|---|---|---|
| MMLU Pro (Artificial Analysis run) | Not measured | 71.3% | Only Llama 3.3 70B Instruct measured |
| SimpleQA Verified | 8.3% | Not measured | Only GPT-4o-mini (2024-07-18) measured |
| MMLU | 76.7% | 86.3% | Llama 3.3 70B Instruct |
| GeoBench | 64.0% | Not measured | Only GPT-4o-mini (2024-07-18) measured |
| LMArena Text: science | 1286 | 1284 | GPT-4o-mini (2024-07-18) |
| LMArena Text: medicine and healthcare | 1282 | 1275 | GPT-4o-mini (2024-07-18) |
| LMArena Text: legal and government | 1305 | 1284 | GPT-4o-mini (2024-07-18) |
| LMArena Text: business and finance | 1265 | 1243 | GPT-4o-mini (2024-07-18) |
| Neither model has a result on the other 12 boards here | |||
Web designNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | GPT-4o-mini (2024-07-18) | Llama 3.3 70B Instruct | Leader |
|---|---|---|---|
| Neither model has a result on 9 boards here | |||
GeneralLike-for-likePublic boards, shown as published, kept out of the index152
| Benchmark | GPT-4o-mini (2024-07-18) | Llama 3.3 70B Instruct | Leader |
|---|---|---|---|
| IFBench | Not measured | 47.1% | Only Llama 3.3 70B Instruct measured |
| Long context reasoning (AA-LCR) | Not measured | 15.7% | Only Llama 3.3 70B Instruct measured |
| DTBench | Not measured | 59.5% | Only Llama 3.3 70B Instruct measured |
| LMCA | Not measured | 17.5% | Only Llama 3.3 70B Instruct measured |
| Fiction.LiveBench | Not measured | 33.3% | Only Llama 3.3 70B Instruct measured |
| Creative writing (Lech Mazur) | 67.2% | Not measured | Only GPT-4o-mini (2024-07-18) measured |
| LMArena Text | 1286 | 1274 | GPT-4o-mini (2024-07-18) |
| LMArena Text: instruction following | 1258 | 1242 | GPT-4o-mini (2024-07-18) |
| LMArena Text: creative writing | 1268 | 1250 | GPT-4o-mini (2024-07-18) |
| LMArena Vision | 1066 | Not measured | Only GPT-4o-mini (2024-07-18) measured |
| LMArena Text: English | 1301 | 1303 | Llama 3.3 70B Instruct |
| LMArena Text: non-English | 1266 | 1236 | GPT-4o-mini (2024-07-18) |
| LMArena Text: Chinese | 1265 | 1217 | GPT-4o-mini (2024-07-18) |
| LMArena Text: German | 1272 | 1251 | GPT-4o-mini (2024-07-18) |
| LMArena Text: Japanese | 1216 | 1150 | GPT-4o-mini (2024-07-18) |
| LMArena Text: Korean | 1195 | 1142 | GPT-4o-mini (2024-07-18) |
| LMArena Text: French | 1297 | 1282 | GPT-4o-mini (2024-07-18) |
| LMArena Text: Spanish | 1276 | 1269 | GPT-4o-mini (2024-07-18) |
| LMArena Text: Russian | 1275 | 1252 | GPT-4o-mini (2024-07-18) |
| LMArena Text: Polish | 1244 | 1259 | Llama 3.3 70B Instruct |
| LMArena Text: multi-turn | 1285 | 1280 | GPT-4o-mini (2024-07-18) |
| LMArena Text: longer queries | 1289 | 1255 | GPT-4o-mini (2024-07-18) |
| LMArena Text: writing and language | 1282 | 1255 | GPT-4o-mini (2024-07-18) |
| LMArena Text: entertainment and media | 1253 | 1241 | GPT-4o-mini (2024-07-18) |
| LMArena Vision: English | 1072 | Not measured | Only GPT-4o-mini (2024-07-18) measured |
| LMArena Vision: Chinese | 1033 | Not measured | Only GPT-4o-mini (2024-07-18) measured |
| Neither model has a result on the other 22 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | GPT-4o-mini (2024-07-18) | Llama 3.3 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-4o-mini (2024-07-18) costs less
Repository review
50K input + 3K output tokens
GPT-4o-mini (2024-07-18) costs less
Agent loop
200K input + 10K output tokens
GPT-4o-mini (2024-07-18) 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-4o-mini (2024-07-18) | Meta Llama 3.3 70B Instruct |
|---|---|---|
| Maker | OpenAI | Meta Llama |
| Weights | No public weights listed | Open weights |
| Released | July 18, 2024 | December 6, 2024 |
| Context window | 128K | 131K |
| Input $/Mtok | $0.15 | $0.45 |
| Output $/Mtok | $0.60 | $0.72 |
| 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-4o-mini (2024-07-18) better than Llama 3.3 70B Instruct for coding?
Neither model has a coding score yet, so the two cannot be ranked on coding. Pick GPT-4o-mini (2024-07-18) for output tokens 1.2x cheaper ($0.60 vs $0.72 per million) and work measured by WeirdML or BALROG, where it leads.
Which is cheaper, GPT-4o-mini (2024-07-18) or Llama 3.3 70B Instruct?
GPT-4o-mini (2024-07-18) costs $0.15 per million input tokens and $0.60 per million output tokens; Llama 3.3 70B Instruct costs $0.45 and $0.72. Cost per solved task is known for neither of them.
Which has the bigger context window, GPT-4o-mini (2024-07-18) or Llama 3.3 70B Instruct?
Llama 3.3 70B Instruct, with 131K tokens against 128K.
How many benchmarks is the GPT-4o-mini (2024-07-18) vs Llama 3.3 70B Instruct comparison based on?
35 benchmarks have a result for both models as of October 22, 2026. GPT-4o-mini (2024-07-18)'s coding index rests on 0 independent benchmarks and Llama 3.3 70B Instruct's on 0. Where only one model has a result, the page says so instead of naming a winner.