OPENAI MODEL GUIDE
GPT-6 Astra API
OpenAI's most capable model for hard end-to-end work across complex reasoning, coding, computer use, research, and document creation.
Choosing and integrating
GPT-6 Astra API pricing breakdown
GPT-6 Astra (gpt-6-astra) is billed by token usage. Prices below are USD per 1M tokens. LLMFly AI rates already include the 0.3x group multiplier; do not apply it again. Provider reference prices are separate. These rates come from the site's Model Plaza snapshot; confirm live catalog and account rates before calling.
Example: a request with 10,000 uncached input tokens and 1,000 output tokens costs approximately $0.045 in LLMFly AI token charges. This excludes cache writes, tools, retries, and other additional fees.
- Platform and provider prices use their own input-token tiers. Do not apply provider long-context multipliers to platform prices.
- Not listed does not mean free. Account for cache reads, cache writes, and uncached input in their respective billing categories.
- Token cost = tokens in each billing category × its price per million ÷ 1,000,000. Add tools and other charges separately.
| Price source | Input tokens per request | Input | Output | Cache write | Cache read | 1h cache write |
|---|---|---|---|---|---|---|
| LLMFly AI | 0 < tokens ≤ 272,000 | $3 | $15 | $3.75 | $0.3 | $3.75 |
| LLMFly AI | 272,000 < tokens | $6 | $22.5 | $7.5 | $0.6 | $7.5 |
| Provider reference | 0 < tokens ≤ 272,000 | $10 | $50 | $12.5 | $1 | $12.5 |
| Provider reference | 272,000 < tokens | $20 | $75 | $25 | $2 | $25 |
Pricing tiers and additional charges
LLMFly AI and vendor reference prices both use a 272K input-token threshold, with the platform applying its 0.3x group multiplier: OpenAI charges the full request at 2x input and cache rates and 1.5x output above 272K input tokens. Batch and Flex are 50% of the vendor Standard rate; Fast mode is 2x the applicable vendor rate. Tools can add per-call charges.
GPT-6 Astra capabilities and workloads
OpenAI's most capable model for hard end-to-end work across complex reasoning, coding, computer use, research, and document creation.
Complex reasoning, software engineering, research, browser automation, document creation, and multi-step professional workflows.
- Streaming, function calling, structured outputs, and reasoning tokens.
- Responses API tools include web search, file search, image generation, code interpreter, hosted shell, apply patch, skills, computer use, MCP, and tool search.
- Supports async tool calling, mid-turn steering, configuration updates, prompt caching, persisted reasoning, and compaction.
- The published knowledge cutoff is April 30, 2026.
Limits to check before migration
- The none reasoning effort is unsupported; use low as the lightest setting.
- For tool calling, OpenAI recommends the Responses API. Chat Completions support does not establish equivalent tool behavior.
- temperature, top_p, and top_logprobs are unsupported; Chat Completions also does not support logprobs.
- Fast mode is unavailable with EU data residency and does not include a latency SLA.
- Fine-tuning is not supported.
Comparing with GPT-5.6 Sol
Astra's Standard vendor rates are $10 input and $50 output per million tokens, versus $4 and $20 for Sol. OpenAI positions Astra as the stronger end-to-end model, but that does not establish route latency or cost per task. Compare pass rate, output tokens, tool loops, and total cost on the same workload.
Use the exact model ID when connecting
Set the model field to gpt-6-astra; do not infer its spelling from the display name. Configure the API address and key using the access documentation; the snippet only shows the model field, not a complete request.
- Send a minimal text request and inspect the response and usage.
- Test streaming, tool calls, and structured output separately if your application needs them.
- Check retry behavior and spending limits before increasing traffic.
{
"model": "gpt-6-astra"
}Vendor capability versus gateway access
GPT-6 Astra is available through the LLMFly AI API. The tools listed above use the Responses API; Fast mode, beta parameters, and regional data policies remain subject to OpenAI's conditions for those features.
Frequently asked questions
What should I put in the model field?
Use the exact ID gpt-6-astra, not the display name.
How is the discount calculated?
The current 0.3x multiplier is applied to the base rate, giving 70% off the catalog's official reference. Amounts are shown per million tokens.
Do these prices include every possible fee?
Not necessarily. Check cache-write duration, tools, and other route charges separately. Use the full pricing table and actual usage records to reconcile a bill.
What are the context and output limits?
GPT-6 Astra is documented with 1,050,000 context and 128,000 maximum output.
Use the GPT-6 Astra API
Use the exact model ID gpt-6-astra and check the current platform rates in Model Plaza.