MiMo V2.5 Pro vs Nemotron 3.5 Lightning
Wins 6 of 6 areas
Coding · Agents · Reasoning · Facts · Long documents · Following instructions
Wins 0 of 6 areas
—
MiMo V2.5 Pro is the stronger all-rounder.Nemotron 3.5 Lightning 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 software30MiMo V2.5 Pro3 of 3 tests
- ReasoningHard problems that need careful thinking30MiMo V2.5 Pro3 of 3 tests
- AgentsCarrying out multi-step tasks on its own20MiMo V2.5 Pro2 of 3 tests · 1 tie
- FactsGetting facts right instead of making them up20MiMo V2.5 Pro2 of 2 tests
- Long documentsFinding answers in very long texts10MiMo V2.5 Pro1 of 1 test
- Following instructionsDoing exactly what it is asked10MiMo V2.5 Pro1 of 1 test
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.
Nemotron 3.5 Lightning costs 90% less for the same work.
The biggest differences
The tests each model wins by the widest margin, up to three each. Scores are out of 100.
Where MiMo V2.5 Pro pulls ahead
- Command-line tasks in a real terminalTerminal-Bench 2.1+40.9points ahead
- Fixes real GitHub issues in Python projectsSWE-bench Verified+27.3points ahead
- Very hard expert questions across many subjectsHumanity's Last Exam+25.1points ahead
Where Nemotron 3.5 Lightning pulls ahead
No clear win on a test scored out of 100.
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.
CodingMiMo V2.5 Pro
- Terminal-Bench 2.1MiMo V2.5 Pro by 40.965.224.3+40.9
- SWE-bench VerifiedMiMo V2.5 Pro by 27.378.951.6+27.3
- SciCodeMiMo V2.5 Pro by 18.550.632.1+18.5
AgentsMiMo V2.5 Pro
- GDPValMiMo V2.5 Pro by 24.1316.9+24.1
- τ-Bench V3 · BankingMiMo V2.5 Pro by 19.98.9+1
- Terminal-Bench 4.0tie00.5tie
ReasoningMiMo V2.5 Pro
- Humanity's Last ExamMiMo V2.5 Pro by 25.135.710.6+25.1
- GPQA DiamondMiMo V2.5 Pro by 12.386.674.3+12.3
- CritPtMiMo V2.5 Pro by 440+4
FactsMiMo V2.5 Pro
- AA-Omniscience · Non-hallucinationMiMo V2.5 Pro by 12.975.362.4+12.9
- AA-Omniscience · AccuracyMiMo V2.5 Pro by 822.414.4+8
Following instructionsMiMo V2.5 Pro
- IFBenchMiMo V2.5 Pro by 8.279.971.7+8.2
Other results9 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.
- Artificial Analysis Coding IndexMiMo V2.5 Pro by 33.460.226.8+33.4
- AA Agentic IndexMiMo V2.5 Pro by 2329.16.1+23
- AA-OmniscienceMiMo V2.5 Pro by 213.3-17.7+21
- MMLU-ProNemotron 3.5 Lightning by 13.468.581.9+13.4
- AA IntelligenceMiMo V2.5 Pro by 13.12612.9+13.1
- MMLUMiMo V2.5 Pro by 10.889.478.6+10.8
- WinoGrandeMiMo V2.5 Pro by 5.685.680+5.6
- ARC-ChallengeMiMo V2.5 Pro by 4.597.292.7+4.5
- HellaSwagMiMo V2.5 Pro by 4.289.885.5+4.2
Questions people ask
Which is better, MiMo V2.5 Pro or Nemotron 3.5 Lightning?
MiMo V2.5 Pro wins all six areas where both have results: coding, agents, reasoning, facts, long documents and following instructions. Nemotron 3.5 Lightning wins none, but costs 90% less.
Which is better for coding?
MiMo V2.5 Pro. It wins 3 of the 3 coding tests both models report; Nemotron 3.5 Lightning wins none.
Which is cheaper?
MiMo V2.5 Pro costs $0.67 per million input tokens and $1.85 per million output tokens; Nemotron 3.5 Lightning costs $0.06 and $0.20. That makes Nemotron 3.5 Lightning about 90% 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 13 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 9 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.