Advertise here
#7

GPT-6.2 Astra

OpenAI Release: 2026-11-17 Tested on: 2026-11-17 19:06 openai/gpt-6.2-astra::medium
~2T total (~150B active)MoEClosedEstimated

GPT-6.2 Astra and Sol tie for #1 in our tests →

Summary

GPT-6.2 Astra scores 9.9 on AI BENCHY and ranks #7. It has 10.0 reliability, a 98.6% pass rate, $1.632 total cost, and 6.65s average response time.

What makes GPT-6.2 Astra unique: It is notably fast compared with similar models.

Model facts

Researched on 2026-11-17

Estimated
Parameters
~2T total (~150B active)
Architecture
MoE
Availability
Closed
License
-

Best estimate from public evidence; the vendor did not disclose every value. GPT-6.2 Astra (flagship) is an OpenRouter-listed GPT-6.2 checkpoint. OpenAI does not disclose parameter counts or architecture. Low-confidence catalog estimate carries forward the GPT-6 Astra/Sol family estimate (2000B total, 150B active); it is not a reported specification. OpenRouter requires reasoning and lists low, medium, high, xhigh, and max efforts. Fallback date records this onboarding request date.

Score

9.9

Consistency

9.6

Total Output Tokens

9,593

Total Input Tokens

155,927

Input Price

$8.000 / 1M

Output Price

$40.000 / 1M

Cache Read Price

$0.400 / 1M

Cache Write Price

$10.000 / 1M

Cache prices apply to input tokens. Reads reuse cached prompts; writes store them and can cost extra. Output tokens use the output price.

Tests Correct

Wrong Tests: 1

Attempt pass rate: 98.6%

Flaky tests

1

Flaky tests had mixed outcomes across runs (at least one pass and one fail).

Response Time (avg)

6.65s

Response Time (max): 46.34s

Response Time (total): 153.03s

Hamster playing table tennis

Prompt: Create a detailed SVG illustration of a hamster playing table tennis.

#7 GPT-6.2 Astra

medium
Cost
$0.106
Time
26.0s
Tokens
2,719 tok

Charts

Choose the first model, then click a second model to open a side-by-side page.

Total Output Tokens

Score vs Total Output Tokens

Quick Compare

Category Breakdown

Category Score Consistency Tests Correct
Agentic 10.0 10.0
Anti-AI Tricks 10.0 10.0
Coding 10.0 10.0
Combined 10.0 10.0
Data parsing and extraction 10.0 10.0
Domain specific 8.2 7.2
General Intelligence 10.0 10.0
Instructions following 10.0 10.0
Puzzle Solving 10.0 10.0
Tool Calling 10.0 10.0
Trivia 10.0 10.0

Compared models