OpenAI / Model guide
GPT-6 Astra
OpenAI's model for complex reasoning, coding and computer use, alongside Claude in shared comparisons.
Model facts
- Context window
- 1.05M
- Max output
- 128K
- Input / 1M tokens
- $10
- Output / 1M tokens
- $50
Effort: low, medium, high, xhigh, max. Text and images in; text out. Reviewed 2026-09-25.
Official model documentationExamples to explore
All 0 examplesNo published cases in this collection yet. Explore the Opus 5.5 collection.
API pricing
$10 input · $50 output · $1 cache read per million tokens.
Standard rates up to 272K input tokens. Above 272K, the full request uses 2× input/cache rates and 1.5× output rates. Tools may add fees.
Official pricingReported benchmarks
| Benchmark | GPT-6 Astra |
|---|---|
| Terminal-Bench 4.0Opus 5.5: xhigh effort | 57.9% |
| FrontierCode v1.1Opus 5.5: max effort | 53.3% |
| CursorBench 4.0Opus 5.5: max effort | Not reported |
| GDPval-AA v2.1Opus 5.5: max effort | 1542 Elo |
| AutomationBenchOpus 5.5: max effort | 41.4% |
| Humanity's Last ExamWith tools · Opus 5.5: max effort | 57.2% |
Vendor-reported launch results, September 22, 2026. Opus 5.5 uses max effort except Terminal-Bench (xhigh). Other models' effort, harness and tool settings can differ; see the original methodology. These are not AI IDE List tests and are separate from Anthropic's medium-effort cost charts.
Benchmark source and methodology