Compare · ModelsLive · 2 picked · head to head

GPT-5.5 vs GPT-5.6 Sol

Side by side · benchmarks, pricing, and signals you can act on.

Winner summary

GPT-5.6 Sol wins 19 of 24 shared benchmarks. Leads in reasoning · knowledge · coding.

Category leads
agentic·GPT-5.5reasoning·GPT-5.6 Solknowledge·GPT-5.6 Solcoding·GPT-5.6 Solgeneral·GPT-5.6 Solmath·GPT-5.6 Sol

Head to head · from the data · data as of 2026-11-17

  • Benchmarks · GPT-5.5 leads on 3 of 24 shared benchmarks (APEX-Agents, GPQA diamond, Mirrorcode), GPT-5.6 Sol on 19 (ARC-AGI, ARC-AGI-2, Chess Puzzles, Deepswe, 15 more) · 2 tied
  • Price · GPT-5.5 input $5.00/M vs GPT-5.6 Sol $4.00/M, GPT-5.6 Sol is 20% cheaper (OpenRouter list price, as of 2026-11-17)
  • Cheapest provider · GPT-5.5 at OpenAI $2.50/M input (of 7 offers) as of 2026-11-17 · GPT-5.6 Sol at OpenAI $2.00/M input (of 8 offers) as of 2026-11-17
  • Context · GPT-5.5 and GPT-5.6 Sol both 1.1M tokens
  • Released · GPT-5.5 Apr 24, 2026 vs GPT-5.6 Sol Jul 9, 2026 · 3 months apart
  • Price history · GPT-5.5 OpenRouter input list price had no move of 5% or more in the 90 days to 2026-11-17

Sources · scores from public leaderboards, list prices and provider offers from OpenRouter, price moves from the BenchGecko changelog (5% or more, held in the next daily check), Gecko tests run by BenchGecko.

Hype vs Reality
GPT-5.5
#64 by perf·#11 by attention
UNDERRATED
GPT-5.6 Sol
#36 by perf·#11 by attention
UNDERRATED
Best value
1.7x better value than GPT-5.5
GPT-5.5
3.3 pts/$
$17.50/M
GPT-5.6 Sol
5.5 pts/$
$12.00/M
Vendor risk
OpenAI logo
OpenAI
$840.0B·Tier 1
Medium risk
OpenAI logo
OpenAI
$840.0B·Tier 1
Medium risk
Head to head
GPT-5.5GPT-5.6 Sol
APEX-Agents
GPT-5.5 leads by +3.7
APEX-Agents · evaluates AI agents on complex, multi-step tasks requiring planning, tool use, and autonomous decision-making in realistic environments.
GPT-5.5
55.1
GPT-5.6 Sol
51.4
ARC-AGI
GPT-5.6 Sol leads by +2.5
ARC-AGI · the original Abstraction and Reasoning Corpus, testing whether AI can solve novel visual pattern recognition tasks without memorization.
GPT-5.5
95.0
GPT-5.6 Sol
97.5
ARC-AGI-2
GPT-5.6 Sol leads by +7.5
ARC-AGI-2 · the second iteration of the Abstraction and Reasoning Corpus, testing novel pattern recognition and abstract reasoning without prior training data.
GPT-5.5
85.0
GPT-5.6 Sol
92.5
Chess Puzzles
GPT-5.6 Sol leads by +5.3
Chess Puzzles · tests strategic and tactical reasoning by having models solve chess puzzle positions, evaluating lookahead and pattern recognition abilities.
GPT-5.5
61.1
GPT-5.6 Sol
66.3
Deepswe
GPT-5.6 Sol leads by +5.6
GPT-5.5
67.0
GPT-5.6 Sol
72.7
Dtbench
GPT-5.5
93.3
GPT-5.6 Sol
93.3
Ebr Bench
GPT-5.6 Sol leads by +10.5
GPT-5.5
34.3
GPT-5.6 Sol
44.8
Frontiercode
GPT-5.6 Sol leads by +4.5
GPT-5.5
43.0
GPT-5.6 Sol
47.5
FrontierMath-Tier-4-v2-Private
GPT-5.6 Sol leads by +10.4
GPT-5.5
72.5
GPT-5.6 Sol
82.9
FrontierMath-Tiers-1-3-v2-Private
GPT-5.6 Sol leads by +4.1
GPT-5.5
89.8
GPT-5.6 Sol
93.8
Furniture Assembly
GPT-5.6 Sol leads by +17.9
GPT-5.5
20.2
GPT-5.6 Sol
38.1
GPQA diamond
GPT-5.5 leads by +0.7
Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs.
GPT-5.5
95.8
GPT-5.6 Sol
95.1
GSO-Bench
GPT-5.6 Sol leads by +36.3
GSO-Bench · evaluates AI models on real-world open-source software engineering tasks, testing the ability to understand and resolve actual GitHub issues.
GPT-5.5
40.2
GPT-5.6 Sol
76.5
Lmca
GPT-5.6 Sol leads by +5.8
GPT-5.5
63.9
GPT-5.6 Sol
69.7
Mirrorcode
GPT-5.5 leads by +3.3
GPT-5.5
23.3
GPT-5.6 Sol
20.0
Mystery Game Puzzles
GPT-5.6 Sol leads by +2.2
GPT-5.5
51.5
GPT-5.6 Sol
53.7
Osworld 2 0
GPT-5.6 Sol leads by +14.3
GPT-5.5
13.0
GPT-5.6 Sol
27.3
OTIS Mock AIME 2024-2025
OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills.
GPT-5.5
100.0
GPT-5.6 Sol
100.0
PostTrainBench
GPT-5.6 Sol leads by +7.1
GPT-5.5
23.9
GPT-5.6 Sol
31.0
Proofbench
GPT-5.6 Sol leads by +33.0
GPT-5.5
50.0
GPT-5.6 Sol
83.0
SimpleBench
GPT-5.6 Sol leads by +3.2
SimpleBench · tests fundamental reasoning capabilities with straightforward problems designed to expose gaps in basic logical and spatial thinking.
GPT-5.5
62.8
GPT-5.6 Sol
66.0
SimpleQA Verified
GPT-5.6 Sol leads by +6.7
SimpleQA Verified · short factual questions with verified answers, measuring factual accuracy and the tendency to hallucinate or provide incorrect information.
GPT-5.5
63.0
GPT-5.6 Sol
69.7
Surface Evolver Bench
GPT-5.6 Sol leads by +5.0
GPT-5.5
88.1
GPT-5.6 Sol
93.1
WeirdML
GPT-5.6 Sol leads by +4.5
WeirdML · tests models on unusual and adversarial machine learning tasks that require creative problem-solving beyond standard patterns.
GPT-5.5
84.9
GPT-5.6 Sol
89.4
Full benchmark table
BenchmarkGPT-5.5GPT-5.6 Sol
APEX-Agents
APEX-Agents · evaluates AI agents on complex, multi-step tasks requiring planning, tool use, and autonomous decision-making in realistic environments.
55.151.4
ARC-AGI
ARC-AGI · the original Abstraction and Reasoning Corpus, testing whether AI can solve novel visual pattern recognition tasks without memorization.
95.097.5
ARC-AGI-2
ARC-AGI-2 · the second iteration of the Abstraction and Reasoning Corpus, testing novel pattern recognition and abstract reasoning without prior training data.
85.092.5
Chess Puzzles
Chess Puzzles · tests strategic and tactical reasoning by having models solve chess puzzle positions, evaluating lookahead and pattern recognition abilities.
61.166.3
Deepswe
67.072.7
Dtbench
93.393.3
Ebr Bench
34.344.8
Frontiercode
43.047.5
FrontierMath-Tier-4-v2-Private
72.582.9
FrontierMath-Tiers-1-3-v2-Private
89.893.8
Furniture Assembly
20.238.1
GPQA diamond
Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs.
95.895.1
GSO-Bench
GSO-Bench · evaluates AI models on real-world open-source software engineering tasks, testing the ability to understand and resolve actual GitHub issues.
40.276.5
Lmca
63.969.7
Mirrorcode
23.320.0
Mystery Game Puzzles
51.553.7
Osworld 2 0
13.027.3
OTIS Mock AIME 2024-2025
OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills.
100.0100.0
PostTrainBench
23.931.0
Proofbench
50.083.0
SimpleBench
SimpleBench · tests fundamental reasoning capabilities with straightforward problems designed to expose gaps in basic logical and spatial thinking.
62.866.0
SimpleQA Verified
SimpleQA Verified · short factual questions with verified answers, measuring factual accuracy and the tendency to hallucinate or provide incorrect information.
63.069.7
Surface Evolver Bench
88.193.1
WeirdML
WeirdML · tests models on unusual and adversarial machine learning tasks that require creative problem-solving beyond standard patterns.
84.989.4
Pricing · per 1M tokens · projected $/mo at 10M tokens
ModelInputOutputContextProjected $/mo
OpenAI logoGPT-5.5$5.00$30.001.1M tokens (~525 books)$112.50
OpenAI logoGPT-5.6 Sol$4.00$20.001.1M tokens (~525 books)$80.00
GPT-5.5 leads on 3 of 24 shared benchmarks (APEX-Agents, GPQA diamond, Mirrorcode), GPT-5.6 Sol on 19 (ARC-AGI, ARC-AGI-2, Chess Puzzles, Deepswe, 15 more) · 2 tied