Skip to main content
TaskPico

GPT-6 Astra rollout

TaskPico Astra Switch Check

Check whether moving an existing OpenAI workload (GPT-5.6 Sol / Terra / Luna) to GPT-6 Astra is available to your plan, what it would cost per month, and which features you rely on are supported.

What you provide

  • Your current model: gpt-5.6-sol, gpt-5.6-terra or gpt-5.6-luna (gpt-5.6 means Sol).
  • Tokens of one typical request: input and output. Requests over 272K input tokens use the long-context rate.
  • Requests per month.
  • Optional: the features you rely on (function calling, structured outputs, web search, file search, code interpreter, computer use, MCP, image input, fine-tuning).

What you get back

  • Access: the current rollout status by plan and API; it cannot tell whether your own account already has Astra.
  • Cost: monthly cost today versus on Astra at Standard uncached list prices, with the formula and the excluded items.
  • Compatibility: each feature you named, marked supported, unsupported or unknown, plus context-window and max-output checks.

It does not compare model quality, does not model caching or batch pricing, and does not estimate your actual bill. Your assistant interprets the numbers; the tool only computes them.

Data

Snapshot as of 2026-09-05, valid until 2026-09-07T12:55:25Z (refreshed by hand from official OpenAI pages; the rollout changes by the day). Data pack version 001.3.

Example

Synthetic workload: Terra, 20,000 input and 2,000 output tokens per request, 10,000 requests a month, using function calling and structured outputs. The numbers are illustrative.

{
  "status": "ok",
  "task_result": {
    "access": {
      "rollout_status": "phased",
      "eligible_chatgpt_plans": [
        "plus",
        "pro",
        "business",
        "enterprise"
      ],
      "enterprise_default_enabled": false,
      "api_rollout_status": "phased",
      "guaranteed_account_access": "unknown",
      "basis": "OpenAI states a phased rollout to Plus, Pro, Business, Enterprise and the API over the coming days; Enterprise starts disabled until an admin enables it. Plan eligibility is not the same fact as a given account already seeing Astra. (openai_astra_launch, as_of 2026-09-05)"
    },
    "cost": {
      "pricing_scope": "standard_uncached",
      "currency": "USD",
      "current": {
        "model": "gpt-5.6-terra",
        "display_name": "GPT-5.6 Terra",
        "input_rate_per_1m_usd": 2,
        "output_rate_per_1m_usd": 12,
        "long_context_applied": false,
        "per_request_usd": 0.064,
        "monthly_usd": 640
      },
      "astra": {
        "model": "gpt-6-astra",
        "display_name": "GPT-6 Astra",
        "input_rate_per_1m_usd": 10,
        "output_rate_per_1m_usd": 50,
        "long_context_applied": false,
        "per_request_usd": 0.3,
        "monthly_usd": 3000
      },
      "delta_monthly_usd": 2360,
      "delta_pct": 368.75,
      "formula": "per_request_usd = input_tokens_per_request × input_rate / 1,000,000 + output_tokens_per_request × output_rate / 1,000,000; monthly_usd = per_request_usd × requests_per_month",
      "excluded_items": []
    },
    "compatibility": [
      {
        "feature": "function_calling",
        "status": "supported",
        "basis": "Astra model page (openai_astra_model), as_of 2026-09-05"
      },
      {
        "feature": "structured_outputs",
        "status": "supported",
        "basis": "Astra model page (openai_astra_model), as_of 2026-09-05"
      },
      {
        "feature": "max_input",
        "status": "supported",
        "basis": "input_tokens_per_request 20,000 vs Astra max input tokens 922,000"
      },
      {
        "feature": "max_output",
        "status": "supported",
        "basis": "output_tokens_per_request 2,000 vs Astra max output 128,000"
      },
      {
        "feature": "context_window",
        "status": "supported",
        "basis": "input + output tokens per request 22,000 vs Astra context window 1,050,000"
      }
    ]
  },
  "cost_estimate_scope": "full",
  "assumptions": [
    "Costs use Standard API list prices for uncached tokens; interpretation of the numbers is left to your assistant.",
    "Long-context rule: a request with more than 272,000 input tokens is priced at 2x input and 1.5x output for the whole request (applies to both models).",
    "Workloads with mixed request sizes should be run as separate scenarios; this check models one request shape."
  ],
  "missing_fields": [],
  "sources": [
    {
      "url": "https://openai.com/index/gpt-6-astra/",
      "as_of": "2026-09-05"
    },
    {
      "url": "https://developers.openai.com/api/docs/models/gpt-6-astra",
      "as_of": "2026-09-05T12:55:22Z"
    },
    {
      "url": "https://developers.openai.com/api/docs/models",
      "as_of": "2026-09-05T12:55:25Z"
    },
    {
      "url": "https://developers.openai.com/api/docs/models/gpt-5.6-sol",
      "as_of": "2026-09-05T12:55:23Z"
    },
    {
      "url": "https://developers.openai.com/api/docs/models/gpt-5.6-terra",
      "as_of": "2026-09-05T12:55:24Z"
    },
    {
      "url": "https://developers.openai.com/api/docs/models/gpt-5.6-luna",
      "as_of": "2026-09-05T12:55:25Z"
    }
  ],
  "capability_version": "1.0.0",
  "data_pack_version": "001.3",
  "as_of": "2026-09-05",
  "valid_until": "2026-09-07T12:55:25Z",
  "next_step": "Review the three sections; re-run with a second request shape if your workload mixes short and long requests."
}
Check my switch to Astra

Use it inside ChatGPT or Claude after connecting TaskPico.

TaskPico Astra Switch Check | TaskPico