Compare · ModelsLive · 2 picked · head to head
gpt-oss-120b vs Llama 3.3 70B Instruct
Side by side · benchmarks, pricing, and signals you can act on.
Winner summary
gpt-oss-120b wins on 8/8 benchmarks
gpt-oss-120b wins 8 of 8 shared benchmarks. Leads in arena · general · knowledge.
Category leads
arena·gpt-oss-120bgeneral·gpt-oss-120bknowledge·gpt-oss-120bmath·gpt-oss-120breasoning·gpt-oss-120bcoding·gpt-oss-120b
Head to head · from the data · data as of 2026-11-15
- Benchmarks · gpt-oss-120b leads on 8 of 8 shared benchmarks (Chatbot Arena Elo · Overall, Dtbench, Fiction.LiveBench, GPQA diamond, 4 more), Llama 3.3 70B Instruct on 0
- Price · gpt-oss-120b input $0.15/M vs Llama 3.3 70B Instruct $0.71/M, gpt-oss-120b is 79% cheaper (OpenRouter list price, as of 2026-11-15)
- Cheapest provider · gpt-oss-120b at AkashML and 4 other providers $0.03/M input (of 26 offers) as of 2026-11-15 · Llama 3.3 70B Instruct at DeepInfra $0.10/M input (of 9 offers) as of 2026-11-15
- Context · gpt-oss-120b and Llama 3.3 70B Instruct both 131K tokens
- Released · gpt-oss-120b Aug 5, 2025 vs Llama 3.3 70B Instruct Dec 6, 2024 · 8 months apart
- Open weights · gpt-oss-120b and Llama 3.3 70B Instruct both have open weights
Sources · scores from public leaderboards, list prices and provider offers from OpenRouter, price moves from the BenchGecko changelog (5% or more, held in the next daily check), Gecko tests run by BenchGecko.
Hype vs Reality
Attention vs performance
gpt-oss-120b
#141 by perf·no signal
Llama 3.3 70B Instruct
#230 by perf·#17 by attention
Best value
gpt-oss-120b
2.6x better value than Llama 3.3 70B Instruct
gpt-oss-120b
131.5 pts/$
$0.38/M
Llama 3.3 70B Instruct
51.0 pts/$
$0.71/M
Vendor risk
Who is behind the model
OpenAI
$840.0B·Tier 1
Meta AI
$1.65T·Tier 1
Head to head
8 benchmarks · 2 models
gpt-oss-120bLlama 3.3 70B Instruct
Chatbot Arena Elo · Overall
gpt-oss-120b leads by +33.8
gpt-oss-120b
1351.6
Llama 3.3 70B Instruct
1317.8
Dtbench
gpt-oss-120b leads by +28.0
gpt-oss-120b
60.5
Llama 3.3 70B Instruct
32.5
Fiction.LiveBench
gpt-oss-120b leads by +11.1
Fiction.LiveBench · a continuously updated benchmark using recently published fiction to test reading comprehension and reasoning, preventing data contamination.
gpt-oss-120b
44.4
Llama 3.3 70B Instruct
33.3
GPQA diamond
gpt-oss-120b leads by +36.5
Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs.
gpt-oss-120b
67.7
Llama 3.3 70B Instruct
31.2
Lmca
gpt-oss-120b leads by +5.4
gpt-oss-120b
26.1
Llama 3.3 70B Instruct
20.6
OTIS Mock AIME 2024-2025
gpt-oss-120b leads by +83.8
OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills.
gpt-oss-120b
88.9
Llama 3.3 70B Instruct
5.0
SimpleBench
gpt-oss-120b leads by +2.6
SimpleBench · tests fundamental reasoning capabilities with straightforward problems designed to expose gaps in basic logical and spatial thinking.
gpt-oss-120b
6.5
Llama 3.3 70B Instruct
3.9
WeirdML
gpt-oss-120b leads by +33.7
WeirdML · tests models on unusual and adversarial machine learning tasks that require creative problem-solving beyond standard patterns.
gpt-oss-120b
48.2
Llama 3.3 70B Instruct
14.4
Full benchmark table
| Benchmark | gpt-oss-120b | Llama 3.3 70B Instruct |
|---|---|---|
Chatbot Arena Elo · Overall | 1351.6 | 1317.8 |
Dtbench | 60.5 | 32.5 |
Fiction.LiveBench Fiction.LiveBench · a continuously updated benchmark using recently published fiction to test reading comprehension and reasoning, preventing data contamination. | 44.4 | 33.3 |
GPQA diamond Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs. | 67.7 | 31.2 |
Lmca | 26.1 | 20.6 |
OTIS Mock AIME 2024-2025 OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills. | 88.9 | 5.0 |
SimpleBench SimpleBench · tests fundamental reasoning capabilities with straightforward problems designed to expose gaps in basic logical and spatial thinking. | 6.5 | 3.9 |
WeirdML WeirdML · tests models on unusual and adversarial machine learning tasks that require creative problem-solving beyond standard patterns. | 48.2 | 14.4 |
Pricing · per 1M tokens · projected $/mo at 10M tokens
| Model | Input | Output | Context | Projected $/mo |
|---|---|---|---|---|
| $0.15 | $0.60 | 131K tokens (~66 books) | $2.62 | |
| $0.71 | $0.71 | 131K tokens (~66 books) | $7.10 |
gpt-oss-120b vs Llama 3.3 70B Instruct · questions answered from the data
gpt-oss-120b leads on 8 of 8 shared benchmarks (Chatbot Arena Elo · Overall, Dtbench, Fiction.LiveBench, GPQA diamond, 4 more), Llama 3.3 70B Instruct on 0