Model comparison
GPT-5.3 Codex vs Qwen2-72B
GPT-5.3 Codex is the stronger model overall, scoring 45.5 to 29.5 on the Noometry Index.
Last verified . 3 shared benchmarks.
Alibaba (Qwen)Summary
- They share 3 benchmarks with published results for both. GPT-5.3 Codex scores higher in 2 categories and Qwen2-72B in 0 categories; 2 gaps are clear of the uncertainty.
- The widest gap is in agentic & tool use, where GPT-5.3 Codex leads 47.8 to 16.9.
- The biggest single-benchmark swing is WeirdML: 79.3% for GPT-5.3 Codex and 11.3% for Qwen2-72B.
- Qwen2-72B has downloadable open weights; the other is API-only.
Side by side
| GPT-5.3 Codex | Qwen2-72B | |
|---|---|---|
| Provider | OpenAI | Alibaba (Qwen) |
| Noometry Index | 45.5 | 29.5 |
| Released | 2026-02-05 | 2024-06-07 |
| Weights | Proprietary | Open |
| Context window | 400K | — |
| Max output | 128K | — |
| Input $ / M tokens | $1.75 | — |
| Output $ / M tokens | $14 | — |
| Results tracked | 8 | 26 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5.3 Codex leads
GPT-5.3 Codex: 48.5 (#51), Qwen2-72B: 28.7 (#303)
| Benchmark | GPT-5.3 Codex | Qwen2-72B |
|---|---|---|
| WeirdML | 79.3% | 11.3% |
| SWE-bench Verified | 74.8% | — |
| LMArena WebDev | 1408 | — |
| BigCodeBench Instruct | — | 38.5% |
| LMArena Coding | — | 1196 |
| BigCodeBench Complete | — | 54% |
| ALE-Bench | 1,655 | — |
Agentic & Tool Use GPT-5.3 Codex leads
GPT-5.3 Codex: 47.8 (#9), Qwen2-72B: 16.9 (#138)
| Benchmark | GPT-5.3 Codex | Qwen2-72B |
|---|---|---|
| METR Time Horizons | 74.5% | 29.9% |
| Terminal-Bench | 78.4% | — |
| TheAgentCompany | — | 1.1% |
| Vending-Bench 2 | 5,940 | — |
Reasoning Not comparable
GPT-5.3 Codex: —, Qwen2-72B: 22.2 (#184)
| Benchmark | GPT-5.3 Codex | Qwen2-72B |
|---|---|---|
| Epoch Capabilities Index | 156.84 | 125.27 |
| LMArena Hard Prompts | — | 1191 |
Math Not comparable
GPT-5.3 Codex: —, Qwen2-72B: 27.9 (#232)
| Benchmark | GPT-5.3 Codex | Qwen2-72B |
|---|---|---|
| LMArena Math | — | 1235 |
| MATH Level 5 | — | 39.1% |
Knowledge Not comparable
GPT-5.3 Codex: —, Qwen2-72B: 21.2 (#270)
| Benchmark | GPT-5.3 Codex | Qwen2-72B |
|---|---|---|
| GPQA Diamond | — | 40.8% |
| LMArena Expert | — | 1171 |
| MMLU | — | 82.4% |
Multilingual Not comparable
GPT-5.3 Codex: —, Qwen2-72B: 36.0 (#239)
| Benchmark | GPT-5.3 Codex | Qwen2-72B |
|---|---|---|
| LMArena Non-English | — | 1176 |
| LMArena Chinese | — | 1240 |
| LMArena French | — | 1169 |
| LMArena German | — | 1151 |
| LMArena Japanese | — | 1111 |
| LMArena Korean | — | 1083 |
| LMArena Russian | — | 1169 |
| LMArena Spanish | — | 1168 |
Instruction Following Not comparable
GPT-5.3 Codex: —, Qwen2-72B: 62.0 (#236)
| Benchmark | GPT-5.3 Codex | Qwen2-72B |
|---|---|---|
| LMArena Instruction Following | — | 1181 |
Long Context Not comparable
GPT-5.3 Codex: —, Qwen2-72B: 36.1 (#230)
| Benchmark | GPT-5.3 Codex | Qwen2-72B |
|---|---|---|
| LMArena Longer Query | — | 1193 |
Writing & Preference Not comparable
GPT-5.3 Codex: —, Qwen2-72B: 41.0 (#236)
| Benchmark | GPT-5.3 Codex | Qwen2-72B |
|---|---|---|
| LMArena Text | — | 1203 |
| LMArena Creative Writing | — | 1181 |
| LMArena Multi-Turn | — | 1195 |
Frequently asked questions
Is GPT-5.3 Codex better than Qwen2-72B?
GPT-5.3 Codex is the stronger model overall, scoring 45.5 to 29.5 on the Noometry Index.
Is GPT-5.3 Codex or Qwen2-72B better for coding?
GPT-5.3 Codex scores higher on coding benchmarks: 48.5 versus 28.7 in the Noometry coding category.
How many benchmarks do GPT-5.3 Codex and Qwen2-72B share?
3 benchmarks have published results for both models. GPT-5.3 Codex has 8 scored results on Noometry and Qwen2-72B has 26.