Gemini 4 Argon vs GPT-6 Astra
Different efforts: high · medium · Compare all at high
Humanity's Last Exam Humanity's Last Exam: Very hard expert questions across many fields, written to stump today’s models
Gemini 4 Argon57.1
GPT-6 Astra52.7
Arena Text · overall Arena Text · overall: Human pairwise preferences in the stated text vote pool
Gemini 4 Argon1,524
GPT-6 Astra · max1,477
Arena Text · Chinese Arena Text · chinese: Human pairwise preferences in the stated text vote pool
Gemini 4 Argon1,608.3
GPT-6 Astra · max1,493
Arena Text · English Arena Text · english: Human pairwise preferences in the stated text vote pool
Gemini 4 Argon1,537.7
GPT-6 Astra · max1,447.9
Arena Text · instruction following Arena Text · instruction following: Human pairwise preferences in the stated text vote pool
Gemini 4 Argon1,535.6
GPT-6 Astra · max1,454.7
Arena Text · non-English Arena Text · non english: Human pairwise preferences in the stated text vote pool
Gemini 4 Argon1,521.7
GPT-6 Astra · max1,432.5
Arena Text · Russian Arena Text · russian: Human pairwise preferences in the stated text vote pool
Gemini 4 Argon1,534
GPT-6 Astra · max1,437
Arena Text · Spanish Arena Text · spanish: Human pairwise preferences in the stated text vote pool
Gemini 4 Argon1,504.4
GPT-6 Astra · max1,422.6
AA Omniscience Index AA Omniscience Index: Scores factual knowledge while penalizing incorrect answers to encourage admitting uncertainty
Gemini 4 Argon42.4
GPT-6 Astra42.2
SimpleQA Verified v1 SimpleQA Verified v1: Accuracy on short factual questions with verified answers, answered without web access
Gemini 4 Argon76
GPT-6 Astra75.8
CritPt (physics) CritPt (physics): Research-level physics problems that need many careful reasoning steps
Gemini 4 Argon27.1
GPT-6 Astra29.1
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 4 Argon52.6
GPT-6 Astra49.6
AA Omniscience · Accuracy AA Omniscience · Accuracy: Measures correct answers on the AA Omniscience factual knowledge test
Gemini 4 Argon49.9
GPT-6 Astra60.6
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 4 Argon15.1
GPT-6 Astra46.5
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 4 Argon79.7
GPT-6 Astra79.7
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 4 Argon21.8
GPT-6 Astra30.4
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 4 Argon99.5
GPT-6 Astra98.8
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 4 Argon98.5
GPT-6 Astra97.5
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 4 Argon97.1
GPT-6 Astra96.7
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 4 Argon91.3
GPT-6 Astra92.1
ARC-AGI-3 · semi-private ARC-AGI-3 · semi-private: Measures interactive game solving through relative human action efficiency
Gemini 4 Argon · low49.8
GPT-6 Astra38.6
Chess Puzzles Chess Puzzles: Models choose the best next move from a novel chess position
Gemini 4 Argon59
GPT-6 Astra · max74
FrontierMath Tier 4 (v2) FrontierMath Tier 4 (v2): Models solve expert-written advanced mathematics problems on the named tier and edition
Gemini 4 Argon78
GPT-6 Astra97.6
FrontierMath Tiers 1–3 (v2) FrontierMath Tiers 1–3 (v2): Models solve expert-written advanced mathematics problems on the named tier and edition
Gemini 4 Argon86
GPT-6 Astra · max93.7
Mystery Game Puzzles Mystery Game Puzzles: Models choose the best move in puzzle positions from an undisclosed game
Gemini 4 Argon64
GPT-6 Astra · max84
MathArena · ArXivMath · August 2026 MathArena · arxiv · 08/2026: Answer accuracy on the named mathematics competition under the recorded effort
Gemini 4 Argon82.5
GPT-6 Astra · max88.6
MathArena · BrokenArXiv · August 2026 MathArena · arxiv_false · 08/2026: Answer accuracy on the named mathematics competition under the recorded effort
Gemini 4 Argon54.2
GPT-6 Astra · max81.9
MathArena · ArXivLean · September 2026 MathArena · arxivlean · 09/2026: Answer accuracy on the named mathematics competition under the recorded effort
Gemini 4 Argon39.6
GPT-6 Astra · max37.7
Vals MedCode Vals MedCode: Medical coding accuracy on patient encounters
Gemini 4 Argon58.8
GPT-6 Astra48.5
Vals MedScribe Vals MedScribe: Accuracy of clinical notes produced from consultation transcripts
Gemini 4 Argon87.4
GPT-6 Astra87.9
Vals Legal Research Bench Vals Legal Research Bench: Grounded legal research; a task passes only if all criteria pass
Gemini 4 Argon54.8
GPT-6 Astra39.4
Vals Science Index Vals Science Index: Scientific problem solving across the index’s component evaluations
Gemini 4 Argon55.4
GPT-6 Astra65.1
Arena Text · French Arena Text · french: Human pairwise preferences in the stated text vote pool
Gemini 4 Argon—
GPT-6 Astra · max1,444.4
Arena Text · German Arena Text · german: Human pairwise preferences in the stated text vote pool
Gemini 4 Argon—
GPT-6 Astra · max1,457.1
Arena Text · Japanese Arena Text · japanese: Human pairwise preferences in the stated text vote pool
Gemini 4 Argon—
GPT-6 Astra · max1,392.3
Arena Text · Korean Arena Text · korean: Human pairwise preferences in the stated text vote pool
Gemini 4 Argon—
GPT-6 Astra · max1,429.6
Arena Text · Polish Arena Text · polish: Human pairwise preferences in the stated text vote pool
Gemini 4 Argon—
GPT-6 Astra · max1,419.5
MLCR-AA MLCR-AA: Models synthesize long medical records; the headline credits concise answers judged both accurate and complete
Gemini 4 Argon—
GPT-6 Astra · max35
FrontierMath Erdos FrontierMath Erdos: Models solve expert-written advanced mathematics problems on the named tier and edition
Gemini 4 Argon—
GPT-6 Astra · max2.9
MathArena · ArXivMath · April 2026 MathArena · arxiv · 04/2026: Answer accuracy on the named mathematics competition under the recorded effort
Gemini 4 Argon—
GPT-6 Astra · max90.8
MathArena · ArXivMath · June 2026 MathArena · arxiv · 06/2026: Answer accuracy on the named mathematics competition under the recorded effort
Gemini 4 Argon—
GPT-6 Astra · max94.4
MathArena · ArXivMath · May 2026 MathArena · arxiv · 05/2026: Answer accuracy on the named mathematics competition under the recorded effort
Gemini 4 Argon—
GPT-6 Astra · max95
MathArena · ArXivMath · September 2026 MathArena · arxiv · 09/2026: Answer accuracy on the named mathematics competition under the recorded effort
Gemini 4 Argon87.7
GPT-6 Astra—
MathArena · BrokenArXiv · April 2026 MathArena · arxiv_false · 04/2026: Answer accuracy on the named mathematics competition under the recorded effort
Gemini 4 Argon—
GPT-6 Astra · max97.5
MathArena · BrokenArXiv · June 2026 MathArena · arxiv_false · 06/2026: Answer accuracy on the named mathematics competition under the recorded effort
Gemini 4 Argon—
GPT-6 Astra · max99.1
MathArena · BrokenArXiv · May 2026 MathArena · arxiv_false · 05/2026: Answer accuracy on the named mathematics competition under the recorded effort
Gemini 4 Argon—
GPT-6 Astra · max91
MathArena · BrokenArXiv · September 2026 MathArena · arxiv_false · 09/2026: Answer accuracy on the named mathematics competition under the recorded effort
Gemini 4 Argon49.7
GPT-6 Astra—
MathArena · ArXivLean · June 2026 MathArena · arxivlean · 06/2026: Answer accuracy on the named mathematics competition under the recorded effort
Gemini 4 Argon—
GPT-6 Astra · max65.2
Vals LegalBench Vals LegalBench: Legal reasoning and legal-language understanding on LegalBench
Gemini 4 Argon88.3
GPT-6 Astra—
| Benchmark | Gemini 4 Argon | GPT-6 Astra |
|---|---|---|
| Humanity's Last Exam | 57.1 | 52.7 |
| Arena Text · overall | 1,524 | 1,477 · max |
| Arena Text · Chinese | 1,608.3 | 1,493 · max |
| Arena Text · English | 1,537.7 | 1,447.9 · max |
| Arena Text · instruction following | 1,535.6 | 1,454.7 · max |
| Arena Text · non-English | 1,521.7 | 1,432.5 · max |
| Arena Text · Russian | 1,534 | 1,437 · max |
| Arena Text · Spanish | 1,504.4 | 1,422.6 · max |
| AA Omniscience Index | 42.4 | 42.2 |
| SimpleQA Verified v1 | 76 | 75.8 |
| CritPt (physics) | 27.1 | 29.1 |
| AA Intelligence Index 4.3.2 | 52.6 | 49.6 |
| AA Omniscience · Accuracy | 49.9 | 60.6 |
| AA Omniscience · Hallucination Rate | 15.1 | 46.5 |
| AA-LCR (long context) 1.1 | 79.7 | 79.7 |
| GDP.pdf · AA all-pass | 21.8 | 30.4 |
| ARC-AGI-1 · public evaluation | 99.5 | 98.8 |
| ARC-AGI-1 · semi-private | 98.5 | 97.5 |
| ARC-AGI-2 · public evaluation | 97.1 | 96.7 |
| ARC-AGI-2 · semi-private | 91.3 | 92.1 |
| ARC-AGI-3 · semi-private | 49.8 · low | 38.6 |
| Chess Puzzles | 59 | 74 · max |
| FrontierMath Tier 4 (v2) | 78 | 97.6 |
| FrontierMath Tiers 1–3 (v2) | 86 | 93.7 · max |
| Mystery Game Puzzles | 64 | 84 · max |
| MathArena · ArXivMath · August 2026 | 82.5 | 88.6 · max |
| MathArena · BrokenArXiv · August 2026 | 54.2 | 81.9 · max |
| MathArena · ArXivLean · September 2026 | 39.6 | 37.7 · max |
| Vals MedCode | 58.8 | 48.5 |
| Vals MedScribe | 87.4 | 87.9 |
| Vals Legal Research Bench | 54.8 | 39.4 |
| Vals Science Index | 55.4 | 65.1 |
| Arena Text · French | — | 1,444.4 · max |
| Arena Text · German | — | 1,457.1 · max |
| Arena Text · Japanese | — | 1,392.3 · max |
| Arena Text · Korean | — | 1,429.6 · max |
| Arena Text · Polish | — | 1,419.5 · max |
| MLCR-AA | — | 35 · max |
| FrontierMath Erdos | — | 2.9 · max |
| MathArena · ArXivMath · April 2026 | — | 90.8 · max |
| MathArena · ArXivMath · June 2026 | — | 94.4 · max |
| MathArena · ArXivMath · May 2026 | — | 95 · max |
| MathArena · ArXivMath · September 2026 | 87.7 | — |
| MathArena · BrokenArXiv · April 2026 | — | 97.5 · max |
| MathArena · BrokenArXiv · June 2026 | — | 99.1 · max |
| MathArena · BrokenArXiv · May 2026 | — | 91 · max |
| MathArena · BrokenArXiv · September 2026 | 49.7 | — |
| MathArena · ArXivLean · June 2026 | — | 65.2 · max |
| Vals LegalBench | 88.3 | — |
Specs
| Spec | Gemini 4 Argon | GPT-6 Astra |
|---|---|---|
| Input, per 1M | !$2 | $10 |
| Output, per 1M | !$10 | $50 |
| Cost per task? | $1.99 | $1.54 |
| Speed, tok/s | 84 | 59 |
| Context | 1M | 1.05M |
| Weights | Closed | Closed |
| Released | Oct 2026 | Sep 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 4 Argon → GPT-6 Astra
Monitoring can return HTTP 403›
Misalignment monitoring can return HTTP 403 before streaming or errors after output; match misalignment_policy_violation and handle stream errors.
No none reasoning effort›
The none reasoning effort is unsupported on GPT-6 Astra and GPT-6.1 Sol; use low instead.
Remove unsupported sampling parameters›
When reasoning effort is not none, remove temperature, top_p and top_logprobs, plus Chat Completions logprobs or Responses output logprobs inclusion.
GPT-6 Astra → Gemini 4 Argon
Context: 1.05M → 1M
Synchronous output stops at 65,536›
With background false, output limits above 65,536 tokens return HTTP 400; stay at or below that limit.
Sampling parameters are ignored›
temperature, top_p and top_k are ignored; remove them from requests.
No explicit caching›
Explicit caching is unavailable; omit explicit cache requests.