GPT-6 Astra Pricing: OpenAI Opens Flagship Model to Enterprises
OpenAI has begun distributing GPT-6 Astra to paying enterprise customers. The model is offered through the Trusted Access Program, and the gpt-6-astra API has been live since September 3. On the standard tier, OpenAI charges $10 for each million input tokens and $50 for each million output tokens. Availability for the Plus, Pro, Business and Enterprise tiers is expected within days.
OpenAI positions Astra for research, coding and extended multi-step work. The model is not aimed at conversational chat. It accepts up to 1.05 million tokens of context and can generate up to 128,000 tokens per response. It is designed for software-agent tasks, such as controlling a browser or another application, and records a 72.6 percent score on the OSWorld benchmark. Access is phased, with Daybreak program companies first. Amazon Bedrock on AWS will also carry the model.
What GPT-6 Astra pricing covers
OpenAI's API documentation shows three rate tiers for gpt-6-astra. Batch and Flex requests are billed at 50 percent below standard rates.
| Tier | Input per 1M tokens | Output per 1M tokens |
|---|---|---|
| Standard (short context) | $10.00 | $50.00 |
| Long context | $20.00 | $75.00 |
| Fast mode | $20.00 | $100.00 |
Astra's cache pricing is separate. Reading cached input costs $1 per million tokens, and writing to cache costs $12.50 per million tokens. Those rates matter for agent loops that continuously revisit the same context.
The API pricing page shows GPT-5.6 Sol next to Astra. OpenAI's current promotional rates for Sol are $4 per million input tokens and $20 per million output tokens. Standard-tier Astra thus costs about 2.5 times as much. The gap is the clearest signal that OpenAI is pricing agent-capable models above their predecessors as Astra exits limited preview. Teams can reduce the bill by half by choosing Batch or Flex processing.
Administrators must enable the model in each workspace before it can be used; it comes switched off by default. Fast mode, priced at double the standard rate, is intended for low-latency calls and becomes available when Plus, Pro, Business and Enterprise access opens in the coming days.
Why this matters
GPT-6 Astra pricing tells enterprise buyers that OpenAI has put agentic computer use at the high end of its lineup. The real cost depends on workload design; cache reuse, batch discounts and context length can each move the bill below the headline rate. Distribution is controlled too, with the model restricted to vetted enterprises until OpenAI widens access.
Sources
ChatGPT — Release Notes | OpenAI Help Center
GPT-6 Astra Model | OpenAI API
GPT-6 Astra: A new generation of intelligence | OpenAI
Related Articles
- GPT-5.6 on Amazon Bedrock: OpenAI's Sol, Terra, and Luna Models Arrive for Enterprise Deployments
- GPT-5.6 Government Review: OpenAI Launches Sol, Terra, Luna
- GPT-5.6 Sol Ultrafast: inference speed as a priced product
✔Human Verified
Researched and cross-referenced against primary sources by the Bytevyte editorial team. This article was generated with the assistance of artificial intelligence and reviewed by the Bytevyte editorial team.