Model comparison
GPT-6 Astra vs Qwen3.8 27B
GPT-6 Astra is the stronger model overall, scoring 70.8 to 46.0 on the Noometry Index. Qwen3.8 27B costs 18× less per token, which makes it the better buy when GPT-6 Astra's lead doesn't matter for your workload.
Last verified . 30 shared benchmarks.
Alibaba (Qwen)Summary
- They share 30 benchmarks with published results for both. GPT-6 Astra scores higher in 9 categories and Qwen3.8 27B in 1 category; 7 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-6 Astra leads 93.5 to 37.1.
- The biggest single-benchmark swing is ProofBench: 99% for GPT-6 Astra and 16% for Qwen3.8 27B.
- Qwen3.8 27B is cheaper at $0.99 / $1.49 per million input/output tokens, against $10 / $50 for GPT-6 Astra.
- GPT-6 Astra accepts more context: 1.05M tokens versus 262K.
- Qwen3.8 27B has downloadable open weights; the other is API-only.
Side by side
| GPT-6 Astra | Qwen3.8 27B | |
|---|---|---|
| Provider | OpenAI | Alibaba (Qwen) |
| Noometry Index | 70.8 | 46.0 |
| Released | 2026-09-03 | 2026-08-14 |
| Weights | Proprietary | Open |
| Context window | 1.05M | 262K |
| Max output | 128K | 33K |
| Input $ / M tokens | $10 | $0.99 |
| Output $ / M tokens | $50 | $1.49 |
| Results tracked | 56 | 31 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-6 Astra leads
GPT-6 Astra: 73.7 (#2), Qwen3.8 27B: 50.5 (#44)
| Benchmark | GPT-6 Astra | Qwen3.8 27B |
|---|---|---|
| LMArena WebDev | 1786 | 1593 |
| SciCode | 56.5% | 46.6% |
| LMArena Coding | 1487 | 1482 |
| DeepSWE | 74.1% | — |
| FrontierCode | 53.3% | — |
| FrontierSWE | 65.5% | — |
| GSO | 79.4% | — |
| WeirdML | 93.6% | — |
| MirrorCode | 46.7% | — |
| ALE-Bench | 2,951 | — |
Agentic & Tool Use GPT-6 Astra leads
GPT-6 Astra: 52.9 (#3), Qwen3.8 27B: 32.9 (#57)
| Benchmark | GPT-6 Astra | Qwen3.8 27B |
|---|---|---|
| APEX-Agents | 64.7% | 47.5% |
| Remote Labor Index | 20.8% | — |
| BALROG | 68.3% | — |
| GDP.pdf | 34.2% | — |
| Vending-Bench 2 | 15,515 | — |
Reasoning GPT-6 Astra leads
GPT-6 Astra: 85.1 (#1), Qwen3.8 27B: 41.0 (#54)
| Benchmark | GPT-6 Astra | Qwen3.8 27B |
|---|---|---|
| ARC-AGI-2 | 95% | 42.4% |
| NYT Connections (extended) | 98.1% | 54.5% |
| ARC-AGI-1 | 98.5% | 87.5% |
| CritPt | 31.7% | 5.4% |
| LMArena Hard Prompts | 1462 | 1460 |
| DTBench | 97.3% | 88% |
| LMCA | 64.4% | 41.4% |
| Epoch Capabilities Index | 166.45 | 149.38 |
| Chess Puzzles | 72% | — |
| EBR-Bench | 76.2% | — |
| Mystery Game Puzzles | 84% | — |
| Surface Evolver Bench | — | 45% |
| Bench to the Future 3 | 0.14 | — |
Math GPT-6 Astra leads
GPT-6 Astra: 93.5 (#2), Qwen3.8 27B: 37.1 (#161)
| Benchmark | GPT-6 Astra | Qwen3.8 27B |
|---|---|---|
| ProofBench | 99% | 16% |
| LMArena Math | 1465 | 1456 |
| FrontierMath (Tiers 1-3) | 93.7% | — |
| FrontierMath Tier 4 | 97.6% | — |
| OTIS Mock AIME 2024-2025 | 100% | — |
| FrontierMath Erdős | 2.9% | — |
Knowledge GPT-6 Astra leads
GPT-6 Astra: 75.3 (#1), Qwen3.8 27B: 41.6 (#109)
| Benchmark | GPT-6 Astra | Qwen3.8 27B |
|---|---|---|
| LMArena Expert | 1483 | 1482 |
| GPQA Diamond | 95.8% | — |
| Humanity's Last Exam | 54.8% | — |
| SimpleQA Verified | 75.6% | — |
| Vectara Hallucination Rate | 8.7% | — |
Multimodal GPT-6 Astra leads
GPT-6 Astra: 55.0 (#3), Qwen3.8 27B: 41.3 (#37)
| Benchmark | GPT-6 Astra | Qwen3.8 27B |
|---|---|---|
| LMArena Vision | 1281 | 1271 |
| Blueprint-Bench 2 | 49.7% | — |
| Furniture Assembly | 80% | — |
| LMArena Document | 1468 | — |
Multilingual Too close to call
GPT-6 Astra: 53.7 (#61), Qwen3.8 27B: 53.7 (#60)
| Benchmark | GPT-6 Astra | Qwen3.8 27B |
|---|---|---|
| LMArena Non-English | 1430 | 1430 |
| LMArena Chinese | 1484 | 1504 |
| LMArena French | 1456 | 1465 |
| LMArena German | 1440 | 1438 |
| LMArena Japanese | 1379 | 1384 |
| LMArena Korean | 1426 | 1393 |
| LMArena Russian | 1436 | 1415 |
| LMArena Spanish | 1407 | 1448 |
Instruction Following Too close to call
GPT-6 Astra: 76.3 (#44), Qwen3.8 27B: 75.8 (#53)
| Benchmark | GPT-6 Astra | Qwen3.8 27B |
|---|---|---|
| LMArena Instruction Following | 1450 | 1439 |
Long Context Too close to call
GPT-6 Astra: 44.5 (#62), Qwen3.8 27B: 44.3 (#70)
| Benchmark | GPT-6 Astra | Qwen3.8 27B |
|---|---|---|
| LMArena Longer Query | 1456 | 1450 |
Writing & Preference GPT-6 Astra leads
GPT-6 Astra: 75.3 (#7), Qwen3.8 27B: 65.8 (#43)
| Benchmark | GPT-6 Astra | Qwen3.8 27B |
|---|---|---|
| LMArena Text | 1441 | 1441 |
| LMArena Creative Writing | 1418 | 1384 |
| EQ-Bench Creative Writing | 2173 | 1671 |
| LMArena Multi-Turn | 1448 | 1441 |
Frequently asked questions
Is GPT-6 Astra better than Qwen3.8 27B?
GPT-6 Astra is the stronger model overall, scoring 70.8 to 46.0 on the Noometry Index. Qwen3.8 27B costs 18× less per token, which makes it the better buy when GPT-6 Astra's lead doesn't matter for your workload.
Which is cheaper, GPT-6 Astra or Qwen3.8 27B?
Qwen3.8 27B is cheaper. It lists at $0.99 per million input tokens and $1.49 per million output tokens; GPT-6 Astra lists at $10 and $50.
Is GPT-6 Astra or Qwen3.8 27B better for coding?
GPT-6 Astra scores higher on coding benchmarks: 73.7 versus 50.5 in the Noometry coding category.
Which has the bigger context window?
GPT-6 Astra does, with 1.05M tokens against 262K.
How many benchmarks do GPT-6 Astra and Qwen3.8 27B share?
30 benchmarks have published results for both models. GPT-6 Astra has 56 scored results on Noometry and Qwen3.8 27B has 31.