Model comparison
GPT-5.3 Codex vs Qwen Turbo
GPT-5.3 Codex is the stronger model overall, scoring 45.8 to 27.1 on the Noometry Index. Qwen Turbo costs 55× less per token, which makes it the better buy when GPT-5.3 Codex's lead doesn't matter for your workload.
Last verified . 0 shared benchmarks.
Alibaba (Qwen)Summary
- Qwen Turbo is cheaper at $0.05 / $0.20 per million input/output tokens, against $1.75 / $14 for GPT-5.3 Codex.
- Qwen Turbo accepts more context: 1M tokens versus 400K.
Side by side
| GPT-5.3 Codex | Qwen Turbo | |
|---|---|---|
| Provider | OpenAI | Alibaba (Qwen) |
| Noometry Index | 45.8 | 27.1 |
| Released | 2026-02-05 | 2024-11-01 |
| Weights | Proprietary | Proprietary |
| Context window | 400K | 1M |
| Max output | 128K | 16K |
| Input $ / M tokens | $1.75 | $0.05 |
| Output $ / M tokens | $14 | $0.20 |
| Results tracked | 8 | 3 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Not comparable
GPT-5.3 Codex: 48.6 (#56), Qwen Turbo: —
| Benchmark | GPT-5.3 Codex | Qwen Turbo |
|---|---|---|
| SWE-bench Verified | 74.8% | — |
| LMArena WebDev | 1409 | — |
| WeirdML | 79.3% | — |
| ALE-Bench | 1,655 | — |
Agentic & Tool Use Not comparable
GPT-5.3 Codex: 48.0 (#9), Qwen Turbo: —
| Benchmark | GPT-5.3 Codex | Qwen Turbo |
|---|---|---|
| Terminal-Bench | 78.4% | — |
| METR Time Horizons | 74.5% | — |
| Vending-Bench 2 | 5,940 | — |
Reasoning Not comparable
GPT-5.3 Codex: —, Qwen Turbo: —
| Benchmark | GPT-5.3 Codex | Qwen Turbo |
|---|---|---|
| Epoch Capabilities Index | 156.77 | — |
Math Not comparable
GPT-5.3 Codex: —, Qwen Turbo: 15.3 (#297)
| Benchmark | GPT-5.3 Codex | Qwen Turbo |
|---|---|---|
| OTIS Mock AIME 2024-2025 | — | 6.1% |
| MATH Level 5 | — | 56.2% |
Knowledge Not comparable
GPT-5.3 Codex: —, Qwen Turbo: 22.2 (#272)
| Benchmark | GPT-5.3 Codex | Qwen Turbo |
|---|---|---|
| GPQA Diamond | — | 41.8% |
Frequently asked questions
Is GPT-5.3 Codex better than Qwen Turbo?
GPT-5.3 Codex is the stronger model overall, scoring 45.8 to 27.1 on the Noometry Index. Qwen Turbo costs 55× less per token, which makes it the better buy when GPT-5.3 Codex's lead doesn't matter for your workload.
Which is cheaper, GPT-5.3 Codex or Qwen Turbo?
Qwen Turbo is cheaper. It lists at $0.05 per million input tokens and $0.20 per million output tokens; GPT-5.3 Codex lists at $1.75 and $14.
Which has the bigger context window?
Qwen Turbo does, with 1M tokens against 400K.
How many benchmarks do GPT-5.3 Codex and Qwen Turbo share?
0 benchmarks have published results for both models. GPT-5.3 Codex has 8 scored results on Noometry and Qwen Turbo has 3.