Skip to content

GPT-6.2 Astra · Math

GPT-6.2 Astra for math: our test runs

GPT-6.2 Astra ran our 5 math prompts in llmwise, through the same pipeline your messages take. Here's every reply as it came, whether it passed, what it cost and what it counts as on Pro, next to the other models.

Based on 5 of our test runs on , through OpenRouter with the app's own prompt and settings.

Short answer

In our test runs on November 17, 2026, GPT-6.2 Astra passed 5 of 5 of our math prompts (both hard ones): tied 1st of 27 models. Each reply counted as 1 message on Pro, where it has 31 a month, and cost $0.0075 to run on average.

Prompt by prompt

Each of the 5 math prompts, and how GPT-6.2 Astra's reply did.

GPT-6.2 Astra on each math prompt
PromptResultCostTime
A discount, then sales taxEverydayPassed: Final answer 66: right.$0.00642.1 s
Pens at 3 for $4EverydayPassed: Final answer 3 bundles and 1 single for 13.50.: right.$0.00762.9 s
Compound interest over three yearsEverydayPassed: Final answer 315.25: right.$0.00782.3 s
Four-digit numbers whose digits sum to 9HardPassed: Final answer 165: right.$0.00762.6 s
The highest of three dice is a 5HardPassed: Final answer 61/216: right.$0.00793.0 s

Every math reply from GPT-6.2 Astra

  1. A discount, then sales tax

    An everyday prompt: “A discount, then sales tax”, in full.

    Passed: Final answer 66: right.

    After the 25% discount: $80 × 0.75 = $60.  
    With 10% sales tax: $60 × 1.10 = $66.
    
    Final answer: $66

    574 tokens in, 46 out · 2.1 s · $0.0064 · 1 message on Pro · answered by openai/gpt-6.2-astra via OpenAI ·

  2. Pens at 3 for $4

    An everyday prompt: “Pens at 3 for $4”, in full.

    Passed: Final answer 3 bundles and 1 single for 13.50.: right.

    Three singles cost $4.50, so each $4 bundle saves $0.50. Use the maximum of 3 bundles for 9 pens, then buy 1 single: \(3 \times \$4 + \$1.50 = \$13.50\).
    
    Final answer: 3 bundles and 1 single for $13.50.

    576 tokens in, 75 out · 2.9 s · $0.0076 · 1 message on Pro · answered by openai/gpt-6.2-astra via OpenAI ·

  3. Compound interest over three years

    An everyday prompt: “Compound interest over three years”, in full.

    Passed: Final answer 315.25: right.

    After 3 years, the balance is \(\$2{,}000 \times 1.05^3 = \$2{,}315.25\). Subtracting the original deposit gives \(\$2{,}315.25 - \$2{,}000 = \$315.25\) in interest.
    
    Final answer: $315.25

    576 tokens in, 80 out · 2.3 s · $0.0078 · 1 message on Pro · answered by openai/gpt-6.2-astra via OpenAI ·

  4. Four-digit numbers whose digits sum to 9

    A hard prompt: “Four-digit numbers whose digits sum to 9”, in full.

    Passed: Final answer 165: right.

    Subtract 1 from the thousands digit, which must be at least 1. The four adjusted digits are now nonnegative and sum to 8, so none can exceed 8.
    
    By stars and bars, the number of ordered solutions is \(\binom{8+3}{3}=\binom{11}{3}=165\).
    
    Final answer: 165

    552 tokens in, 79 out · 2.6 s · $0.0076 · 1 message on Pro · answered by openai/gpt-6.2-astra via OpenAI ·

  5. The highest of three dice is a 5

    A hard prompt: “The highest of three dice is a 5”, in full.

    Passed: Final answer 61/216: right.

    There are \(6^3=216\) equally likely outcomes. Of these, \(5^3\) have all dice showing 5 or less. Subtracting the \(4^3\) outcomes where all dice show 4 or less leaves \(5^3-4^3=61\) with a highest roll of exactly 5.
    
    Final answer: 61/216

    567 tokens in, 84 out · 3.0 s · $0.0079 · 1 message on Pro · answered by openai/gpt-6.2-astra via OpenAI ·

How it compares on math

GPT-6.2 Astra beside the job's picks among every model. 27 models share the hard-problems pick, GPT-6.2 Astra among them.

GPT-6.2 Astra beside other models on math
ModelPassedOn ProCost per reply
GPT-6.1 TerraOur pick: hard problems (shared), best value5 of 5250 a month on Pro$0.0005
GPT-6 LunaOur pick: hard problems (shared), everyday5 of 560 a day on Pro$0.0001
GPT-6.2 AstraOur pick: hard problems (shared)5 of 531 a month on Pro$0.0075
On Pro: Pro's count on each model. Cost: what OpenRouter charged us per reply, on average.

Every limit is published. Paid plans also have a monthly fair-use limit on AI cost: Plus $3.75, Pro $7.50, Max $20, Ultra $42, Studio $85. Using every message on your plan at typical sizes stays under it; very large messages and heavy research use it faster. Every limit, explained.

GPT-6.2 Astra next to each model on math

GPT-6.2 Astra against each model on the same 5 math prompts. One run each, through OpenRouter: a wait depends on the provider and the load that day, so a lead under 10% counts as close.

  • Against Claude Fable 5.2: GPT-6.2 Astra answered within 10% of its time (2.6 s to 2.8 s), cost 1.2× more, and passed 5 of 5 to its 5.

  • Against Claude Fable 5.1: GPT-6.2 Astra answered 1.5× later (2.6 s to 1.7 s), cost 1.2× less, and passed 5 of 5 to its 5.

  • Against Claude Opus 5.5: GPT-6.2 Astra answered within 10% of its time (2.6 s to 2.7 s), cost 1.7× more, and passed 5 of 5 to its 5.

  • Against Claude Sonnet 5.5: GPT-6.2 Astra answered 1.5× later (2.6 s to 1.8 s), cost 4.3× more, and passed 5 of 5 to its 5.

  • Against Claude Sonnet 5: GPT-6.2 Astra answered within 10% of its time (2.6 s to 2.5 s), cost 1.8× more, and passed 5 of 5 to its 5.

  • Against Claude Haiku 5.5: GPT-6.2 Astra answered 1.7× later (2.6 s to 1.6 s), cost 50.0× more, and passed 5 of 5 to its 5.

  • Against Claude Haiku 4.5: GPT-6.2 Astra answered 1.3× later (2.6 s to 2.1 s), cost 4.3× more, and passed 5 of 5 to its 5.

  • Against GPT-6 Astra: GPT-6.2 Astra answered 1.5× sooner (2.6 s to 3.9 s), cost 1.5× more, and passed 5 of 5 to its 5.

  • Against GPT-6.2 Sol: GPT-6.2 Astra answered 1.3× later (2.6 s to 2.0 s), cost 4.0× more, and passed 5 of 5 to its 5.

  • Against GPT-6.1 Sol: GPT-6.2 Astra answered 4.1× sooner (2.6 s to 10.9 s), cost 7.2× more, and passed 5 of 5 to its 5.

  • Against GPT-6 Sol: GPT-6.2 Astra answered within 10% of its time (2.6 s to 2.5 s), cost 6.8× more, and passed 5 of 5 to its 5.

  • Against GPT-6.1 Terra: GPT-6.2 Astra answered 1.2× later (2.6 s to 2.2 s), cost 14.6× more, and passed 5 of 5 to its 5.

  • Against GPT-6 Luna: GPT-6.2 Astra answered 1.2× sooner (2.6 s to 3.3 s), cost 142.2× more, and passed 5 of 5 to its 5.

  • Against Gemini 4 Carbon: GPT-6.2 Astra answered 2.3× sooner (2.6 s to 6.1 s), cost 3.0× more, and passed 5 of 5 to its 5.

  • Against Gemini 4 Argon: GPT-6.2 Astra answered 2.7× sooner (2.6 s to 7.1 s), cost 5.1× more, and passed 5 of 5 to its 5.

  • Against Gemini 4 Hydrogen: GPT-6.2 Astra answered 1.1× sooner (2.6 s to 2.9 s), cost 3.3× more, and passed 5 of 5 to its 5.

  • Against Gemini 3.1 Pro (preview): GPT-6.2 Astra answered 2.5× sooner (2.6 s to 6.5 s), cost 1.6× more, and passed 5 of 5 to its 5.

  • Against Gemini 3.8 Flash: GPT-6.2 Astra answered 1.5× sooner (2.6 s to 4.0 s), cost 8.3× more, and passed 5 of 5 to its 5.

  • Against DeepSeek V4.1 Flash: GPT-6.2 Astra answered 3.7× later (2.6 s to 0.7 s), cost 25.6× more, and passed 5 of 5 to its 5.

  • Against DeepSeek V4 Pro: GPT-6.2 Astra answered 1.7× sooner (2.6 s to 4.4 s), cost 5.1× more, and passed 5 of 5 to its 5.

  • Against Grok 4.7: GPT-6.2 Astra answered 3.3× sooner (2.6 s to 8.8 s), cost within 10% of it, and passed 5 of 5 to its 5.

  • Against Kimi K3.1 Pro: GPT-6.2 Astra answered 1.3× sooner (2.6 s to 3.4 s), cost 2.6× more, and passed 5 of 5 to its 5.

  • Against Kimi K3: GPT-6.2 Astra answered 2.0× sooner (2.6 s to 5.2 s), cost 2.0× more, and passed 5 of 5 to its 5.

  • Against GLM 5.3: GPT-6.2 Astra answered 2.2× later (2.6 s to 1.2 s), cost 10.0× more, and passed 5 of 5 to its 5.

  • Against GLM 5.3 Flash: GPT-6.2 Astra answered 2.6× later (2.6 s to 1.0 s), cost 45.2× more, and passed 5 of 5 to its 5.

  • Against Mistral Large 4: GPT-6.2 Astra answered 2.4× sooner (2.6 s to 6.4 s), cost 6.8× more, and passed 5 of 5 to its 5.

How these runs were done

Final answer. Automatic. The reply's last “Final answer:” line must hold the right value.

How the runs were done, and every math prompt.

GPT-6.2 Astra, and math, elsewhere

Questions

Is GPT-6.2 Astra good for math?

In our test runs it passed 5 of 5 math prompts, tied 1st of the 27 models in llmwise. Every reply is on this page, so you can judge them yourself.

How many of my messages does a math reply from GPT-6.2 Astra use?

1 message each on Pro, where it has 31 a month on Pro. The price of a message is fixed and shown before you send it, however long the reply.

How were these runs done?

The same way for every model: each prompt sent through llmwise's own pipeline, each reply checked the same way. The methods page has every prompt and how each is scored.

Claude, GPT, Gemini, DeepSeek, Grok, Kimi, GLM, and Mistral, in one chat.

See what a message costs before you send it. Every plan is month to month: cancel in two clicks. Sign in with an email link, no password.