Gemini 3.1 Pro vs Solar Pro4
Wins 4 of 6 areas
Coding · Reasoning · Math · Long documents
Wins 0 of 6 areas
—
Gemini 3.1 Pro is the stronger all-rounder.Solar Pro4 is cheaper.
Scores updated · 22 tests both models report · How we compare
Where each one wins
Tests won in each of the six areas where both have results. Each piece is one test, so a longer bar means more evidence; grey means the two scored within a point of each other.
- CodingWriting and fixing software40Gemini 3.1 Pro4 of 4 tests
- ReasoningHard problems that need careful thinking30Gemini 3.1 Pro3 of 3 tests
- MathCompetition and research-level math10Gemini 3.1 Pro1 of 1 test
- Long documentsFinding answers in very long texts10Gemini 3.1 Pro1 of 1 test
- AgentsCarrying out multi-step tasks on its own22Even2 each
- FactsGetting facts right instead of making them up11Even1 each
Math and long documents 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.
Solar Pro4 costs 89% less for the same work.
The biggest differences
The three tests each model wins by the widest margin. Scores are out of 100.
Where Gemini 3.1 Pro pulls ahead
- Finds hard-to-locate facts by browsing the webBrowseComp+36.7points ahead
- Answers hard knowledge questions correctlyAA-Omniscience · Accuracy+36points ahead
- Very hard expert questions across many subjectsHumanity's Last Exam+17.8points ahead
Where Solar Pro4 pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+26.5points ahead
- Real work tasks from 44 professionsGDPVal+15.8points ahead
- Customer-service tasks in a simulated bankτ-Bench V3 · Banking+1.9points ahead
Every test, side by side
All 22 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingGemini 3.1 Pro
- Terminal-Bench 2.1Gemini 3.1 Pro by 16.573.857.3+16.5
- SciCodeGemini 3.1 Pro by 14.158.744.6+14.1
- SWE-bench VerifiedGemini 3.1 Pro by 1080.670.6+10
- LMArena · WebDevGemini 3.1 Pro by 76 rating points14471371+76 rating
AgentsEven
- BrowseCompGemini 3.1 Pro by 36.785.949.2+36.7
- GDPValSolar Pro4 by 15.814.730.5+15.8
- Terminal-Bench 4.0Gemini 3.1 Pro by 3.540.5+3.5
- τ-Bench V3 · BankingSolar Pro4 by 1.921.423.3+1.9
ReasoningGemini 3.1 Pro
- Humanity's Last ExamGemini 3.1 Pro by 17.84729.2+17.8
- CritPtGemini 3.1 Pro by 12.317.75.4+12.3
- GPQA DiamondGemini 3.1 Pro by 594.189.1+5
FactsEven
- AA-Omniscience · AccuracyGemini 3.1 Pro by 3654.918.9+36
- AA-Omniscience · Non-hallucinationSolar Pro4 by 26.549.175.6+26.5
Other results7 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.
- AA-OmniscienceGemini 3.1 Pro by 32.731.9-0.8+32.7
- AA Agentic IndexSolar Pro4 by 23.310.333.6+23.3
- Artificial Analysis Coding IndexGemini 3.1 Pro by 16.168.852.7+16.1
- τ³-Bench BankingSolar Pro4 by 6.516.523+6.5
- MMLU-ProGemini 3.1 Pro by 4.79186.3+4.7
- LiveCodeBenchGemini 3.1 Pro by 3.991.787.8+3.9
- AA IntelligenceGemini 3.1 Pro by 1.529.728.2+1.5
Questions people ask
Which is better, Gemini 3.1 Pro or Solar Pro4?
Gemini 3.1 Pro wins four of the six areas where both have results: coding, reasoning, math and long documents. Solar Pro4 wins none, but costs 89% less. They are level on agents and facts.
Which is better for coding?
Gemini 3.1 Pro. It wins 4 of the 4 coding tests both models report; Solar Pro4 wins none.
Which is cheaper?
Gemini 3.1 Pro costs $2.00 per million input tokens and $12.00 per million output tokens; Solar Pro4 costs $0.30 and $1.20. That makes Solar Pro4 about 89% cheaper for the same work.
How do you compare the two?
We use the 22 benchmark tests both models have published scores on. The verdict counts the 15 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 7 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.