Anthropic Claude Opus 4.1 vs OpenAI GPT-6 Astra
Side by side from live data: prices, limits, who serves each model and how reliably, and how they do on our chess arena.
In short
- GPT-6 Astra is 1.5× cheaper at a typical mix of 3 input tokens per output token.
- GPT-6 Astra accepts longer prompts: 1M vs 200K tokens.
- GPT-6 Astra is the newer model (released 2026-09-04).
| Claude Opus 4.1 | GPT-6 Astra | |
|---|---|---|
| Vendor | Anthropic | OpenAI |
| Released | 2025-08-05 | 2026-09-04 |
| Input / 1M | $15.00 | $10.00 |
| Output / 1M | $75.00 | $50.00 |
| Cached input / 1M | $1.50 | $1.00 |
| Blended / 1M (3:1) | $30.00 | $20.00 |
| Context | 200K | 1M |
| Max output | 32K | 128K |
| Accepts | text, image, file | text, image, file |
| Reasoning | Yes | Yes |
| Providers on OpenRouter | 1 | 7 |
| Cheapest provider, output | $75.00 | $50.00 |
| Best provider uptime, 24 h | 100% | 100% |
| Chess arena rating | not playing | not playing |
| Vendor API status | Operational | Operational |
Cost calculator
Claude Opus 4.1
$2,025 per month$67.50 per day
GPT-6 Astra
$1,350 per month$45.00 per day
GPT-6 Astra saves $675 a month on this workload.
At list prices per 1M tokens; prompt caching and batch discounts are not included.
Compare Claude Opus 4.1 with