Beam 1.0 vs Gemini 3.1 Pro
Wins 2 of 5 areas
Coding · Following instructions
Wins 1 of 5 areas
Reasoning
Beam 1.0 wins more areas, narrowly.Gemini 3.1 Pro is better at reasoning.
Scores updated · 18 tests both models report · How we compare
Where each one wins
Tests won in each of the five 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 software41Beam 1.04 of 5 tests
- Following instructionsDoing exactly what it is asked10Beam 1.01 of 1 test
- ReasoningHard problems that need careful thinking04Gemini 3.1 Pro4 of 4 tests
- AgentsCarrying out multi-step tasks on its own11Even1 each
- MathCompetition and research-level math00Even0 each · 1 tie
Math 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.
The biggest differences
The three tests each model wins by the widest margin. Scores are out of 100.
Where Beam 1.0 pulls ahead
- Long, multi-file coding tasks in real codebasesSWE-bench Pro+11.9points ahead
- Multi-step tasks using many real software toolsMCP Atlas+11.6points ahead
- Command-line tasks in a real terminalTerminal-Bench 2.1+6.3points ahead
Where Gemini 3.1 Pro pulls ahead
- Abstract visual puzzles that people can solveARC-AGI-2+53.6points ahead
- Very hard expert questions across many subjectsHumanity's Last Exam+11points ahead
- Code for real scientific research problemsSciCode+9.5points ahead
Every test, side by side
All 18 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingBeam 1.0
- SWE-bench ProBeam 1.0 by 11.966.154.2+11.9
- SciCodeGemini 3.1 Pro by 9.549.258.7+9.5
- Terminal-Bench 2.1Beam 1.0 by 6.380.173.8+6.3
- SWE-bench MultilingualBeam 1.0 by 1.478.376.9+1.4
- SWE-bench VerifiedBeam 1.0 by 181.680.6+1
AgentsEven
- MCP AtlasBeam 1.0 by 11.680.869.2+11.6
- BrowseCompGemini 3.1 Pro by 9.176.885.9+9.1
ReasoningGemini 3.1 Pro
- ARC-AGI-2Gemini 3.1 Pro by 53.623.577.1+53.6
- Humanity's Last ExamGemini 3.1 Pro by 113647+11
- GPQA DiamondGemini 3.1 Pro by 3.590.694.1+3.5
- CritPtGemini 3.1 Pro by 1.416.317.7+1.4
Other results5 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 1.1Beam 1.0 by 31.343.312+31.3
- ARC-AGI-1Gemini 3.1 Pro by 247498+24
- τ³-Bench BankingBeam 1.0 by 19.636.116.5+19.6
- DeepSearchQA (F1)Gemini 3.1 Pro by 1.480.581.9+1.4
- HLE (with tools)tie51.151.4tie
Questions people ask
Which is better, Beam 1.0 or Gemini 3.1 Pro?
Beam 1.0 wins two of the five areas where both have results: coding and following instructions. Gemini 3.1 Pro wins reasoning. They are level on agents and math.
Which is better for coding?
Beam 1.0. It wins 4 of the 5 coding tests both models report; Gemini 3.1 Pro wins 1.
How do you compare the two?
We use the 18 benchmark tests both models have published scores on. The verdict counts the 13 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 5 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.