Gemini 3.8 Flash vs GPT-5.6 Sol
Humanity's Last Exam Humanity's Last Exam: Very hard expert questions across many fields, written to stump today’s models
Gemini 3.8 Flash42.1
GPT-5.6 Sol42.2
Arena Text · overall Arena Text · overall: Human pairwise preferences in the stated text vote pool
Gemini 3.8 Flash · high1,496
GPT-5.6 Sol · xhigh1,485
Arena Text · Chinese Arena Text · chinese: Human pairwise preferences in the stated text vote pool
Gemini 3.8 Flash · high1,550.8
GPT-5.6 Sol · xhigh1,526.6
Arena Text · English Arena Text · english: Human pairwise preferences in the stated text vote pool
Gemini 3.8 Flash · high1,499
GPT-5.6 Sol · xhigh1,458
Arena Text · French Arena Text · french: Human pairwise preferences in the stated text vote pool
Gemini 3.8 Flash · high1,491.7
GPT-5.6 Sol · xhigh1,474.5
Arena Text · German Arena Text · german: Human pairwise preferences in the stated text vote pool
Gemini 3.8 Flash · high1,497.6
GPT-5.6 Sol · xhigh1,475.5
Arena Text · instruction following Arena Text · instruction following: Human pairwise preferences in the stated text vote pool
Gemini 3.8 Flash · high1,491
GPT-5.6 Sol · xhigh1,482.3
Arena Text · Japanese Arena Text · japanese: Human pairwise preferences in the stated text vote pool
Gemini 3.8 Flash · high1,506.3
GPT-5.6 Sol · xhigh1,470.9
Arena Text · Korean Arena Text · korean: Human pairwise preferences in the stated text vote pool
Gemini 3.8 Flash · high1,461.6
GPT-5.6 Sol · xhigh1,442.1
Arena Text · non-English Arena Text · non english: Human pairwise preferences in the stated text vote pool
Gemini 3.8 Flash · high1,491.1
GPT-5.6 Sol · xhigh1,450.8
Arena Text · Polish Arena Text · polish: Human pairwise preferences in the stated text vote pool
Gemini 3.8 Flash · high1,498.5
GPT-5.6 Sol · xhigh1,461.4
Arena Text · Russian Arena Text · russian: Human pairwise preferences in the stated text vote pool
Gemini 3.8 Flash · high1,514.4
GPT-5.6 Sol · xhigh1,470.4
Arena Text · Spanish Arena Text · spanish: Human pairwise preferences in the stated text vote pool
Gemini 3.8 Flash · high1,485.7
GPT-5.6 Sol · xhigh1,440.6
AA Omniscience Index AA Omniscience Index: Scores factual knowledge while penalizing incorrect answers to encourage admitting uncertainty
Gemini 3.8 Flash28.6
GPT-5.6 Sol19.4
SimpleQA Verified v1 SimpleQA Verified v1: Accuracy on short factual questions with verified answers, answered without web access
Gemini 3.8 Flash74.6
GPT-5.6 Sol · max69.7
CritPt (physics) CritPt (physics): Research-level physics problems that need many careful reasoning steps
Gemini 3.8 Flash12.3
GPT-5.6 Sol22.9
AA Intelligence Index 4.3.2 AA Intelligence Index 4.3.2: Artificial Analysis combines tests of knowledge, reasoning, coding and tool use into one score
Gemini 3.8 Flash39.8
GPT-5.6 Sol39.2
AA Omniscience · Accuracy AA Omniscience · Accuracy: Measures correct answers on the AA Omniscience factual knowledge test
Gemini 3.8 Flash53
GPT-5.6 Sol57.8
AA Omniscience · Hallucination Rate AA Omniscience · Hallucination Rate: Measures how often attempted answers are incorrect on the AA Omniscience factual knowledge test · lower is better
Gemini 3.8 Flash51.9
GPT-5.6 Sol90.8
AA-LCR (long context) 1.1 AA-LCR (long context) 1.1: Questions that need reading and connecting facts across very long documents (around 100K tokens)
Gemini 3.8 Flash84
GPT-5.6 Sol80.3
GDP.pdf · AA all-pass GDP.pdf · AA all-pass: Models reason over professional PDF documents; an attempt passes only when every atomic criterion passes
Gemini 3.8 Flash22.8
GPT-5.6 Sol26.2
MLCR-AA MLCR-AA: Models synthesize long medical records; the headline credits concise answers judged both accurate and complete
Gemini 3.8 Flash · high21.7
GPT-5.6 Sol15
ARC-AGI-1 · public evaluation ARC-AGI-1 · public-eval: Models infer transformations from examples and solve unseen grid puzzles; the evaluation split is explicit
Gemini 3.8 Flash97.8
GPT-5.6 Sol93.9
ARC-AGI-1 · semi-private ARC-AGI-1 · semi-private: Models infer transformations from examples and solve unseen grid puzzles; the evaluation split is explicit
Gemini 3.8 Flash97.5
GPT-5.6 Sol92.5
ARC-AGI-2 · public evaluation ARC-AGI-2 · public-eval: Models infer transformations from examples and solve unseen grid puzzles; the evaluation split is explicit
Gemini 3.8 Flash87.5
GPT-5.6 Sol70.7
ARC-AGI-2 · semi-private ARC-AGI-2 · semi-private: Models infer transformations from examples and solve unseen grid puzzles; the evaluation split is explicit
Gemini 3.8 Flash82.9
GPT-5.6 Sol67.1
ARC-AGI-3 · semi-private ARC-AGI-3 · semi-private: Measures interactive game solving through relative human action efficiency
Gemini 3.8 Flash4
GPT-5.6 Sol1.1
Chess Puzzles Chess Puzzles: Models choose the best next move from a novel chess position
Gemini 3.8 Flash · high58
GPT-5.6 Sol · max59
FrontierMath Erdos FrontierMath Erdos: Models solve expert-written advanced mathematics problems on the named tier and edition
Gemini 3.8 Flash · high0
GPT-5.6 Sol · max0
FrontierMath Tier 4 (v2) FrontierMath Tier 4 (v2): Models solve expert-written advanced mathematics problems on the named tier and edition
Gemini 3.8 Flash · high22
GPT-5.6 Sol · max82.9
FrontierMath Tiers 1–3 (v2) FrontierMath Tiers 1–3 (v2): Models solve expert-written advanced mathematics problems on the named tier and edition
Gemini 3.8 Flash · high68.4
GPT-5.6 Sol · max89.1
Mystery Game Puzzles Mystery Game Puzzles: Models choose the best move in puzzle positions from an undisclosed game
Gemini 3.8 Flash · high47
GPT-5.6 Sol · max58
MathArena · ArXivMath · June 2026 MathArena · arxiv · 06/2026: Answer accuracy on the named mathematics competition under the recorded effort
Gemini 3.8 Flash67.4
GPT-5.6 Sol · max88.5
MathArena · BrokenArXiv · June 2026 MathArena · arxiv_false · 06/2026: Answer accuracy on the named mathematics competition under the recorded effort
Gemini 3.8 Flash20.8
GPT-5.6 Sol · max67.3
Vals LegalBench Vals LegalBench: Legal reasoning and legal-language understanding on LegalBench
Gemini 3.8 Flash87
GPT-5.6 Sol87
Vals MedCode Vals MedCode: Medical coding accuracy on patient encounters
Gemini 3.8 Flash48.1
GPT-5.6 Sol44
Vals MedScribe Vals MedScribe: Accuracy of clinical notes produced from consultation transcripts
Gemini 3.8 Flash84.5
GPT-5.6 Sol85.2
Vals Legal Research Bench Vals Legal Research Bench: Grounded legal research; a task passes only if all criteria pass
Gemini 3.8 Flash38.9
GPT-5.6 Sol48.1
Vals Science Index Vals Science Index: Scientific problem solving across the index’s component evaluations
Gemini 3.8 Flash35.8
GPT-5.6 Sol41.5
IFBench (instructions) IFBench (instructions): Checks whether the model follows precise formatting and content instructions
Gemini 3.8 Flash—
GPT-5.6 Sol69.6
MathArena · ArXivMath · April 2026 MathArena · arxiv · 04/2026: Answer accuracy on the named mathematics competition under the recorded effort
Gemini 3.8 Flash64.2
GPT-5.6 Sol—
MathArena · ArXivMath · August 2026 MathArena · arxiv · 08/2026: Answer accuracy on the named mathematics competition under the recorded effort
Gemini 3.8 Flash40.4
GPT-5.6 Sol—
MathArena · ArXivMath · May 2026 MathArena · arxiv · 05/2026: Answer accuracy on the named mathematics competition under the recorded effort
Gemini 3.8 Flash58.3
GPT-5.6 Sol—
MathArena · ArXivMath · September 2026 MathArena · arxiv · 09/2026: Answer accuracy on the named mathematics competition under the recorded effort
Gemini 3.8 Flash50
GPT-5.6 Sol—
MathArena · BrokenArXiv · April 2026 MathArena · arxiv_false · 04/2026: Answer accuracy on the named mathematics competition under the recorded effort
Gemini 3.8 Flash33.6
GPT-5.6 Sol—
MathArena · BrokenArXiv · August 2026 MathArena · arxiv_false · 08/2026: Answer accuracy on the named mathematics competition under the recorded effort
Gemini 3.8 Flash27.4
GPT-5.6 Sol—
MathArena · BrokenArXiv · May 2026 MathArena · arxiv_false · 05/2026: Answer accuracy on the named mathematics competition under the recorded effort
Gemini 3.8 Flash19
GPT-5.6 Sol—
MathArena · BrokenArXiv · September 2026 MathArena · arxiv_false · 09/2026: Answer accuracy on the named mathematics competition under the recorded effort
Gemini 3.8 Flash26.7
GPT-5.6 Sol—
MathArena · ArXivLean · June 2026 MathArena · arxivlean · 06/2026: Answer accuracy on the named mathematics competition under the recorded effort
Gemini 3.8 Flash19.6
GPT-5.6 Sol—
| Benchmark | Gemini 3.8 Flash | GPT-5.6 Sol |
|---|---|---|
| Humanity's Last Exam | 42.1 | 42.2 |
| Arena Text · overall | 1,496 · high | 1,485 · xhigh |
| Arena Text · Chinese | 1,550.8 · high | 1,526.6 · xhigh |
| Arena Text · English | 1,499 · high | 1,458 · xhigh |
| Arena Text · French | 1,491.7 · high | 1,474.5 · xhigh |
| Arena Text · German | 1,497.6 · high | 1,475.5 · xhigh |
| Arena Text · instruction following | 1,491 · high | 1,482.3 · xhigh |
| Arena Text · Japanese | 1,506.3 · high | 1,470.9 · xhigh |
| Arena Text · Korean | 1,461.6 · high | 1,442.1 · xhigh |
| Arena Text · non-English | 1,491.1 · high | 1,450.8 · xhigh |
| Arena Text · Polish | 1,498.5 · high | 1,461.4 · xhigh |
| Arena Text · Russian | 1,514.4 · high | 1,470.4 · xhigh |
| Arena Text · Spanish | 1,485.7 · high | 1,440.6 · xhigh |
| AA Omniscience Index | 28.6 | 19.4 |
| SimpleQA Verified v1 | 74.6 | 69.7 · max |
| CritPt (physics) | 12.3 | 22.9 |
| AA Intelligence Index 4.3.2 | 39.8 | 39.2 |
| AA Omniscience · Accuracy | 53 | 57.8 |
| AA Omniscience · Hallucination Rate | 51.9 | 90.8 |
| AA-LCR (long context) 1.1 | 84 | 80.3 |
| GDP.pdf · AA all-pass | 22.8 | 26.2 |
| MLCR-AA | 21.7 · high | 15 |
| ARC-AGI-1 · public evaluation | 97.8 | 93.9 |
| ARC-AGI-1 · semi-private | 97.5 | 92.5 |
| ARC-AGI-2 · public evaluation | 87.5 | 70.7 |
| ARC-AGI-2 · semi-private | 82.9 | 67.1 |
| ARC-AGI-3 · semi-private | 4 | 1.1 |
| Chess Puzzles | 58 · high | 59 · max |
| FrontierMath Erdos | 0 · high | 0 · max |
| FrontierMath Tier 4 (v2) | 22 · high | 82.9 · max |
| FrontierMath Tiers 1–3 (v2) | 68.4 · high | 89.1 · max |
| Mystery Game Puzzles | 47 · high | 58 · max |
| MathArena · ArXivMath · June 2026 | 67.4 | 88.5 · max |
| MathArena · BrokenArXiv · June 2026 | 20.8 | 67.3 · max |
| Vals LegalBench | 87 | 87 |
| Vals MedCode | 48.1 | 44 |
| Vals MedScribe | 84.5 | 85.2 |
| Vals Legal Research Bench | 38.9 | 48.1 |
| Vals Science Index | 35.8 | 41.5 |
| IFBench (instructions) | — | 69.6 |
| MathArena · ArXivMath · April 2026 | 64.2 | — |
| MathArena · ArXivMath · August 2026 | 40.4 | — |
| MathArena · ArXivMath · May 2026 | 58.3 | — |
| MathArena · ArXivMath · September 2026 | 50 | — |
| MathArena · BrokenArXiv · April 2026 | 33.6 | — |
| MathArena · BrokenArXiv · August 2026 | 27.4 | — |
| MathArena · BrokenArXiv · May 2026 | 19 | — |
| MathArena · BrokenArXiv · September 2026 | 26.7 | — |
| MathArena · ArXivLean · June 2026 | 19.6 | — |
Specs
| Spec | Gemini 3.8 Flash | GPT-5.6 Sol |
|---|---|---|
| Input, per 1M | !$0.75 | $4 |
| Output, per 1M | !$3.75 | $20 |
| Cost per task? | $0.93 | $0.50 |
| Speed, tok/s | — | 85 |
| Context | 1M | 1.05M |
| Weights | Closed | Closed |
| Released | Sep 2026 | Jul 2026 |
| See images | ✓ | ✓ |
| Read PDFs | ✓ | Partial |
| Hear audio | ✓ | — |
| Watch video | ✓ | — |
| Call tools | ✓ | ✓ |
| Force a tool call | Partial | Partial |
| Return strict JSON | ✓ | ✓ |
| Search the web | ✓ | ? |
| Run code | ✓ | ? |
| Use a computer | ? | ? |
What breaks when you switch
Gemini 3.8 Flash → GPT-5.6 Sol
Nothing known.
GPT-5.6 Sol → Gemini 3.8 Flash
Context: 1.05M → 1M
Sampling parameters are ignored›
temperature, top_p and top_k are ignored now and will cause HTTP 400 in future generations; remove them from requests.
No minimal thinking level›
The minimal thinking level returns an error; avoid minimal.