Vascend
HorizonVascend IDEVascend CLICommunityBenchmarkContact
Sign inJoin Discord
HomeHorizonIDECLISign in
ChangelogIntelligenceCost per taskCoding Agent IndexCyber IndexDecision IndexBenchmarksSpeedLatencyPriceLeaderboard
Benchmark

GPT-6 Astra (Max)

  • OpenAI
  • Release · September 3, 2026
  • Proprietary
  • Reasoning

Intelligence

Intelligence Index

#9 of 266
  1. 55.856Muse Wonder Full (Max)
  2. 55.155Kimi K3.1 Pro
  3. 53.654Claude Opus 5.5 (High)
  4. 53.453Claude Fable 5.1 (Max)
  5. 53.253Claude Fable 5.1 (Xhigh)
  6. 52.753GPT-6 Astra (Max)
  7. 52.653Gemini 4 Argon (High)
  8. 52.452GPT-6 Astra (Xhigh)
  9. 52.452Muse Wonder (Max)
  10. 51.952Claude Sonnet 5.5 (Xhigh)
Data from Artificial Analysis

Evaluations

  • AA-Briefcase1,568#11
  • AA-Omniscience Accuracy62.6%#10
  • AA-Omniscience Non-Hallucination48.7%#19
  • AutomationBench-AA68.5%#7
  • GDPval-AA1,575#17
  • Humanity's Last Exam54.7%#9
  • GDP.pdf31%#3
  • Terminal-Bench 4.059.1%#6
  • AA-LCR80.7%#37
  • SciCode56.5%#25
  • CritPt31.7%#2
  • Terminal-Bench-Science63.3%#1
  • FrontierCode47%#5
  • AA-AnalystAgent51.3%#7
  • Harvey LAB-AA8.6%#4
  • AA-3D1,164#3
  • ITBench-AA48.6%#5
  • MLCR-AA41.7%#1
  • MMMU-Pro86.9%#2
  • PersonalityBench: Expressiveness54.3#15
  • PersonalityBench: Organization67#14
  • PersonalityBench: Transparency94.9#3
  • PersonalityBench: Slop13.7#22
  • PersonalityBench: Informativeness21.3#23
  • PersonalityBench: Warmth72.1#9
  • PersonalityBench: Conviction41.6#18
  • PersonalityBench: Imagination72.1#8
  • GPQA Diamond96.1%#2
  • τ²-Bench Banking41.4%#14
GPT-6 Astra (Max)Median
Data from Artificial Analysis

Cost per task

Cost per task

$ per Intelligence Index task · #18 of 24
  1. $1.131.13Mistral Large 4 Preview
  2. $1.241.24Gemini 3.8 Flash (High)
  3. $1.991.99Gemini 4 Argon (High)
  4. $2.012.01GLM-5.3 (Max)
  5. $2.112.11Kimi K3.1 Pro
  6. $3.263.26GPT-6 Astra (Max)
  7. $3.743.74Grok 4.7 (Xhigh)
  8. $4.544.54Muse Wonder (Max)
  9. $5.415.41Qwen3.8 Max (0902)
  10. $5.465.46Claude Sonnet 5.5 (Max)
Data from Artificial Analysis

Time per task

seconds per task · #10 of 26
  1. 263263Step 5 Preview
  2. 284284Muse Spark Open (Xhigh)
  3. 341341Muse Wonder (Max)
  4. 364364GPT-6 Luna (Max)
  5. 396396DeepSeek V4.1 Flash 1017 (Max)
  6. 412412GPT-6 Astra (Max)
  7. 428428Claude Haiku 5.5 (Max)
  8. 498498Gemini 3.8 Flash (High)
  9. 507507MiniMax-M3.1-Flash
  10. 578578Mistral Large 4 Preview
Data from Artificial Analysis

Tokens per task

output tokens per task · #2 of 27
  1. 26.3K26KBeam 1.0
  2. 27.2K27KGPT-6 Astra (Max)
  3. 34.7K35KInkling (Xhigh)
  4. 38.1K38KGPT-6.1 Sol (Max)
  5. 40.4K40KMuse Spark Open (Xhigh)
  6. 49.2K49KNemotron 3.5 Super VL
  7. 50K50KGPT-6 Luna (Max)
  8. 51.5K52KK2 Horizon 375B A23B
  9. 53.8K54KMuse Wonder Full (Max)
  10. 54.5K55KMuse Wonder (Max)
Data from Artificial Analysis

Speed and price

Speed

tokens per second · #116 of 162
  1. 7777Ministral 3 14B
  2. 7777Magistral Small 1.2
  3. 7676LongCat 2.0
  4. 7070GLM-5.3-Flash
  5. 7070Claude Fable 5.1 (Max)
  6. 6666GPT-6 Astra (Max)
  7. 6666North Mini Code
  8. 6464Kimi K2.7 Code
  9. 6363Claude Fable 5.1 (Xhigh)
  10. 6262GPT-6 Astra (Xhigh)
Data from Artificial Analysis

Latency

seconds to first token · #159 of 162
  1. 102.7103GPT-5.6 Terra (Max)
  2. 106.6107GPT-6.1 Sol (Xhigh)
  3. 109.6110Claude Opus 5.5 (Xhigh)
  4. 161.5162GPT-6 Astra (Xhigh)
  5. 180.5181Claude Fable 5.1 (Max)
  6. 236236GPT-6.1 Sol (Max)
  7. 278.3278GPT-6 Astra (Max)
  8. 322.2322Claude Haiku 5.5 (Max)
  9. 371.6372Claude Sonnet 5.5 (Max)
  10. 652.6653Claude Opus 5.5 (Max)
Data from Artificial Analysis

Price

$ per 1M tokens (3:1) · #162 of 169
  1. $88Claude Opus 5.5 (Medium)
  2. $88Claude Opus 5.5 (Low)
  3. $11.25$11GPT-5.5 Instant (June 2026)
  4. $2020Claude Fable 5.1 (Max)
  5. $2020Claude Fable 5.1 (Xhigh)
  6. $2020GPT-6 Astra (Max)
  7. $2020GPT-6 Astra (Xhigh)
  8. $2020Claude Fable 5.1 (High)
  9. $2020GPT-6 Astra (High)
  10. $2020GPT-6 Astra (Medium)
Data from Artificial Analysis

Nearby models

  • Muse Wonder Full (Max)Meta55.8
  • Kimi K3.1 ProKimi55.1
  • Claude Opus 5.5 (High)Anthropic53.6
  • Claude Fable 5.1 (Max)Anthropic53.4
  • Claude Fable 5.1 (Xhigh)Anthropic53.2
  • Gemini 4 Argon (High)Google52.6
  • GPT-6 Astra (Xhigh)OpenAI52.4
  • Muse Wonder (Max)Meta52.4
Data from Artificial Analysis

Same lab

  • GPT-6.1 Sol (Max)OpenAI51.8
  • GPT-6.1 Sol (Xhigh)OpenAI51
  • GPT-6 Astra (High)OpenAI50.9
  • GPT-6.1 Sol (High)OpenAI50.2
  • GPT-6 Astra (Medium)OpenAI49.6
  • GPT-6.1 Sol (Medium)OpenAI47.8
  • GPT-6 Astra (Low)OpenAI45.8
  • GPT-5.6 Terra (Max)OpenAI42.1
Data from Artificial Analysis
HorizonVascend IDEOlympusTalosCoursesCommunityBenchmark
DiscordGitHubXLet’s talkLegalPrivacyCookiesData useSecurity
© 2026 Vascend Technologies · VAT IT04911550616
ascend