OpenAI
GPT-6 Astra
ReasoningClosed weights
- Intelligence
- 45.8
- Rank 35 of 744 measured models
- Coding
- 75.7
- Rank 27 of 298 measured models
- Output speed
- 42.3 t/s
- Rank 355 of 403 measured models
- Price/1M
- $20
- Rank 638 of 659 measured models
- Context window
- 1M
- Rank 84 of 1199 measured models
- Cost per task
- $0.818
- Rank 120 of 193 measured models
GPT-6 Astra (Low) scores 45.8 on the Artificial Analysis Intelligence Index, ranking 35th of 744 measured models. The closest models on this index are Grok 4.7 (46.4), MiMo-V2.6-Pro (46.3), and Qwen3.8 Max (0902) (45.4). At $20 per million tokens (3:1 input/output blended price), it is more expensive than Grok 4.7 ($3), MiMo-V2.6-Pro ($0.544), and Qwen3.8 Max (0902) ($3). With an output speed of 42.3 t/s, it is about as fast as Qwen3.8 Max (0902) (41.9 t/s), slower than Grok 4.7 (78.7 t/s) and MiMo-V2.6-Pro (45.8 t/s).
Models at a similar level
The closest models on the Intelligence Index, one configuration per family.
- Grok 4.7
SpaceXSI · $3 · 78.7 t/s
46.4+0.6$378.7 t/sCompare - MiMo-V2.6-Pro
Xiaomi · $0.544 · 45.8 t/s
46.3+0.5$0.54445.8 t/sCompare - Qwen3.8 Max (0902)
Alibaba · $3 · 41.9 t/s
45.4−0.4$341.9 t/sCompare - GLM-5.3
Z.AI · $2.15 · 58.2 t/s
44.8−1.0$2.1558.2 t/sCompare
