OpenAI
GPT-6.2 Astra
OpenAI's new Astra for demanding software, research and professional documents. Bring a detailed brief and a way to verify the result.
Replaces GPT-6 Astra: input $10.5 → $8.4 per 1M (-20%), output $52.5 → $42 per 1M (-20%), same context window.
Run this model
openai/gpt-6.2-astra
Use an endpoint supported by this model.
Get a key for this setup →const response = await fetch('https://api.minirouter.sh/v1/chat/completions', { method: 'POST', headers: { Authorization: `Bearer ${process.env.MINIROUTER_KEY}`, 'Content-Type': 'application/json', }, body: JSON.stringify({ model: 'openai/gpt-6.2-astra', messages: [{ role: 'user', content: 'Why is the sky blue?' }], max_tokens: 1024, }),}) const data = await response.json()console.log(data.choices[0].message.content)Start with text and low reasoning effort. These models can reserve their full 128,000-token output allowance even when you request a short answer; keep enough balance for the hold. Omit temperature and top_p, and do not set reasoning to none.
Model overview
- Input / 1M
- $8.4
- All-in MiniRouter rate
- Output / 1M
- $42
- All-in MiniRouter rate
- Context
- 1.1M
- 1,050,000 tokens
- Max output
- 128K
- tokens
- Reasoning
- low → max
- Five effort settings
- Best fit
- Complex projects
- Code, research and documents
- Announced
- 17 Nov 2026
- publisher release
Compare text-model alternatives or estimate token costs for your workload.
Use Astra where a stronger answer changes the outcome.
- [ 01 ]
Complex code changes
Supply the relevant interfaces, tests and failure symptoms. Ask for a plan that addresses the failure before proposing code.
- [ 02 ]
Document synthesis
Provide source labels, the intended audience and a concrete deliverable. Review every claim that affects a decision.
- [ 03 ]
Research with checks
Separate assumptions, observations and calculations. Ask for reproducible steps that another person can verify.
Make the first result easy to check.
- [ 01 ]
Give it the evidence
Provide the relevant code or extracted document text. Ask for source references where the answer depends on a particular passage.
- [ 02 ]
Set the effort deliberately
The catalog offers low, medium, high, xhigh and max. Start at low, then compare a higher setting only when the task needs it.
- [ 03 ]
Check what came back
Run proposed tests and verify calculations. Validate machine-readable answers in your application before using them.
Keep a baseline you can rerun.
- [ 01 ]
Keep sampling options out
Omit temperature and top_p. Reasoning cannot be disabled with none.
- [ 02 ]
Separate the model from the application
A model response does not run a browser, edit a file or execute a command. Your application supplies those tools and controls their permissions.
- [ 03 ]
Read the current route details
Check the model page before switching protocols or inputs. Schema-enforced output is not currently exposed for these MiniRouter routes.
One route, standard clients.
https://api.minirouter.sh/v1openai/gpt-6.2-astra/v1/chat/completions/v1/messages/v1/responsesConfigured providers. Priority follows the live catalog order.
| Provider | Priority | Input / 1M | Output / 1M | Parameters | Access |
|---|---|---|---|---|---|
vercel | 1 | $8.4 | $42 | max_tokens, stop, tools, tool_choice, reasoning, include_reasoning | MiniRouter balance |