Gemini 3.1 Pro vs GPT-6 Astra
Wins 2 of 8 areas
Long documents · Following instructions
Wins 5 of 8 areas
Coding · Agents · Reasoning · Facts · Math
GPT-6 Astra is the stronger all-rounder.Gemini 3.1 Pro is cheaper and better at long documents.
Both rank among the ten best models we track in facts.
Scores updated · 43 tests both models report · How we compare
Where each one wins
Tests won in each of the eight areas we test. Each piece is one test, so a longer bar means more evidence; grey means the two scored within a point of each other.
- ReasoningHard problems that need careful thinking07GPT-6 Astra7 of 7 tests
- AgentsCarrying out multi-step tasks on its own05GPT-6 Astra5 of 5 tests
- CodingWriting and fixing software14GPT-6 Astra4 of 5 tests
- FactsGetting facts right instead of making them up03GPT-6 Astra3 of 4 tests · 1 tie
- MathCompetition and research-level math03GPT-6 Astra3 of 3 tests
- Long documentsFinding answers in very long texts10Gemini 3.1 Pro1 of 1 test
- Following instructionsDoing exactly what it is asked10Gemini 3.1 Pro1 of 1 test
- Images and chartsUnderstanding pictures, charts and video11Even1 each
Long documents and following instructions rest on a single test each.
What it costs
Prices per million tokens, roughly 750,000 words. The bars show the cost of a million tokens read plus a million written.
Gemini 3.1 Pro costs 77% less for the same work.
The biggest differences
The three tests each model wins by the widest margin. Scores are out of 100.
Where GPT-6 Astra pulls ahead
- Solves new puzzle games it has never seenARC-AGI-3+99.5points ahead
- Extremely hard research-level math problemsFrontierMath Tier 4+70.8points ahead
- Complex command-line tasks across many fieldsTerminal-Bench 4.0+55.1points ahead
Where Gemini 3.1 Pro pulls ahead
- Follows detailed instructions exactlyLiveBench · Instruction Following+3.5points ahead
- Code for real scientific research problemsSciCode+2.2points ahead
- Reasons across sets of long documentsAA-LCR+1.3points ahead
Every test, side by side
All 43 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingGPT-6 Astra
- LMArena · WebDevGPT-6 Astra by 328 rating points14471775+328 rating
- Terminal-Bench 2.1GPT-6 Astra by 14.673.888.4+14.6
- LiveBench · Agentic CodingGPT-6 Astra by 13.244.157.3+13.2
- LiveBench · CodingGPT-6 Astra by 3.976.580.4+3.9
- SciCodeGemini 3.1 Pro by 2.258.756.5+2.2
AgentsGPT-6 Astra
- Terminal-Bench 4.0GPT-6 Astra by 55.1459.1+55.1
- GDPValGPT-6 Astra by 3914.653.6+39
- τ-Bench V3 · BankingGPT-6 Astra by 2021.441.4+20
- AA IT-Bench SREGPT-6 Astra by 18.330.348.6+18.3
- BrowseCompGPT-6 Astra by 5.685.991.5+5.6
ReasoningGPT-6 Astra
- ARC-AGI-3GPT-6 Astra by 99.50.499.9+99.5
- ARC-AGI-2GPT-6 Astra by 17.977.195+17.9
- CritPtGPT-6 Astra by 1417.731.7+14
- LiveBench · ReasoningGPT-6 Astra by 8.78492.7+8.7
- Humanity's Last ExamGPT-6 Astra by 7.74754.7+7.7
- SimpleBenchGPT-6 Astra by 479.683.6+4
- GPQA DiamondGPT-6 Astra by 294.196.1+2
FactsGPT-6 Astra
- AA-Omniscience · AccuracyGPT-6 Astra by 7.754.962.6+7.7
- SimpleQA VerifiedGPT-6 Astra by 2.173.575.6+2.1
- Vectara HHEM hallucination ratelower is betterGPT-6 Astra by 1.710.48.7+1.7
- AA-Omniscience · Non-hallucinationtie49.148.7tie
Images and chartsEven
- MMMU-ProGPT-6 Astra by 4.582.486.9+4.5
- LMArena · VisionGemini 3.1 Pro by 16 rating points12961280+16 rating
MathGPT-6 Astra
- FrontierMath Tier 4GPT-6 Astra by 70.826.897.6+70.8
- FrontierMath Tiers 1-3 (v2)GPT-6 Astra by 3459.693.7+34
- LiveBench · MathematicsGPT-6 Astra by 5.89196.8+5.8
Following instructionsGemini 3.1 Pro
- LiveBench · Instruction FollowingGemini 3.1 Pro by 3.579.175.6+3.5
Other results15 tests, not counted
Tests outside the eight areas. They are not counted above: several are summary scores built from other tests, or the same test under another name.
- DeepSWE 1.1GPT-6 Astra by 62.11274.1+62.1
- AA Agentic IndexGPT-6 Astra by 41.210.351.5+41.2
- HiL-Bench (Tools-allowed)GPT-6 Astra by 3035.365.3+30
- AA IntelligenceGPT-6 Astra by 2329.752.7+23
- PostTrainBenchGPT-6 Astra by 22.721.644.3+22.7
- LVBenchGPT-6 Astra by 21.366.287.5+21.3
- AA-OmniscienceGPT-6 Astra by 11.531.943.4+11.5
- Artificial Analysis Coding IndexGPT-6 Astra by 8.168.876.9+8.1
- HLE (with tools)GPT-6 Astra by 5.851.457.2+5.8
- livebench_data_analysisGPT-6 Astra by 4.578.583+4.5
- vectara_avg_summary_lengthGPT-6 Astra by 40.8 rating points107.7148.5+40.8 rating
- livebench_languageGPT-6 Astra by 485.489.4+4
- vectara_factual_consistencyGPT-6 Astra by 1.789.691.3+1.7
- vectara_answer_ratetie99.4100tie
- ARC-AGI-1tie9897.5tie
Questions people ask
Which is better, Gemini 3.1 Pro or GPT-6 Astra?
GPT-6 Astra wins five of the eight areas we test: coding, agents, reasoning, facts and math. Gemini 3.1 Pro wins long documents and following instructions, and costs 77% less. They are level on images and charts.
Which is better for coding?
GPT-6 Astra. It wins 4 of the 5 coding tests both models report; Gemini 3.1 Pro wins 1.
Which is cheaper?
Gemini 3.1 Pro costs $2.00 per million input tokens and $12.00 per million output tokens; GPT-6 Astra costs $10.00 and $50.00. That makes Gemini 3.1 Pro about 77% cheaper for the same work.
How do you compare the two?
We use the 43 benchmark tests both models have published scores on. The verdict counts the 28 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 15 are listed but not counted, because several are summary scores or repeat a test. Each score is the one shown on the model's own page.