GPT-6.2 Astra vs GPT-6.2 Sol

OpenAI against OpenAI, the same lab's consecutive models. Prices, scores and speed below are live.

  • GPT-6.2 Sol costs 75% less on blended price ($6.00 against $24.00 per 1M tokens, 3 input to 1 output).
  • GPT-6.2 Astra scores higher on the Artificial Analysis Intelligence Index: 62.1 against 57.3.
  • On the 11 benchmarks both have scores on, GPT-6.2 Astra leads on 10 and GPT-6.2 Sol on 1.
  • Both read up to 1.1M tokens of context.
  • GPT-6.2 Astra is newer: listed Nov 17, 2026, against Nov 17, 2026.

Side by side

GPT-6.2 AstraGPT-6.2 Sol
LabOpenAIOpenAI
ListedNov 17, 2026Nov 17, 2026
Input per 1M tokens$12.00$3.00
Output per 1M tokens$60.00$15.00
Cached input per 1M$0.60$0.15
Blended (3:1) per 1M$24.00$6.00
Context window1.1M1.1M
Max output128K128K
Knowledge cutoff--
Intelligence index62.157.3
Coding index--
ReasoningYesYes
Reads imagesYesYes
Tool useYesYes
Structured outputYesYes

Every benchmark both are scored on

Each model's best score per benchmark, with its source. Scores from different sources or settings are not always comparable; each links to where it was published.

BenchmarkGPT-6.2 AstraGPT-6.2 Sol
APEX-Agents72.2%Epoch AI66.6%Epoch AI
ARC-AGI-198.5%Epoch AI97.5%Epoch AI
ARC-AGI-295.8%Epoch AI85.8%Epoch AI
Artificial Analysis Intelligence Index62.1Artificial Analysis and OpenRouter57.3Artificial Analysis and OpenRouter
LMArena (Coding)1450 EloLMArena1440 EloLMArena
LMArena (Overall)1414 EloLMArena1414 EloLMArena
LMArena: Creative Writing1417 EloLMArena1403 EloLMArena
LMArena: Hard Prompts1425 EloLMArena1419 EloLMArena
LMArena: Longer Query1431 EloLMArena1427 EloLMArena
LMArena: Multi-Turn1425 EloLMArena1420 EloLMArena
SimpleBench77.3%Epoch AI69.9%Epoch AI

More comparisons

Where these figures come from

Prices are GenMagic's own, what the studio and the API charge, recomputed from each provider's price list every few minutes. The indices are the Artificial Analysis indices, and every benchmark score links to its source, re-read daily. Measured speed is GenMagic's own generation records, in aggregate. See OpenAI models, OpenAI models, every LLM's price and our sources.