Nemotron 3 Ultra 550B A55B vs Qwen3.8 Flash Next
Wins 1 of 7 areas
Math
Wins 4 of 7 areas
Coding · Agents · Reasoning · Following instructions
Qwen3.8 Flash Next wins more areas, narrowly.Nemotron 3 Ultra 550B A55B is better at math.
Both rank among the ten best models we track in following instructions.
Scores updated · 43 tests both models report · How we compare
Where each one wins
Tests won in each of the seven 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 software08Qwen3.8 Flash Next8 of 8 tests
- AgentsCarrying out multi-step tasks on its own04Qwen3.8 Flash Next4 of 4 tests
- ReasoningHard problems that need careful thinking04Qwen3.8 Flash Next4 of 4 tests
- Following instructionsDoing exactly what it is asked01Qwen3.8 Flash Next1 of 2 tests · 1 tie
- MathCompetition and research-level math10Nemotron 3 Ultra 550B A55B1 of 1 test
- FactsGetting facts right instead of making them up11Even1 each
- Long documentsFinding answers in very long texts00Even0 each · 1 tie
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.
Qwen3.8 Flash Next costs 81% 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 Qwen3.8 Flash Next pulls ahead
- Finds hard-to-locate facts by browsing the webBrowseComp+44.6points ahead
- Command-line tasks in a real terminalTerminal-Bench 2.1+32.2points ahead
- Customer-service tasks in a simulated bankτ-Bench V3 · Banking+31.2points ahead
Where Nemotron 3 Ultra 550B A55B pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+15.6points ahead
- Fresh competition math problemsLiveBench · Mathematics+2.9points 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.
CodingQwen3.8 Flash Next
- Terminal-Bench 2.1Qwen3.8 Flash Next by 32.253.986.1+32.2
- LiveBench · Agentic CodingQwen3.8 Flash Next by 22.938.761.6+22.9
- SWE-bench MultilingualQwen3.8 Flash Next by 13.367.781+13.3
- SWE-bench ProQwen3.8 Flash Next by 12.849.762.5+12.8
- SciCodeQwen3.8 Flash Next by 10.340.350.6+10.3
- SWE-bench VerifiedQwen3.8 Flash Next by 870.778.7+8
- LiveCodeBench v6Qwen3.8 Flash Next by 2.98991.9+2.9
- LiveBench · CodingQwen3.8 Flash Next by 1.970.772.5+1.9
AgentsQwen3.8 Flash Next
- BrowseCompQwen3.8 Flash Next by 44.644.489+44.6
- τ-Bench V3 · BankingQwen3.8 Flash Next by 31.214.245.4+31.2
- GDPValQwen3.8 Flash Next by 30.725.756.4+30.7
- Terminal-Bench 4.0Qwen3.8 Flash Next by 24.80.525.3+24.8
ReasoningQwen3.8 Flash Next
- LiveBench · ReasoningQwen3.8 Flash Next by 12.774.787.4+12.7
- Humanity's Last ExamQwen3.8 Flash Next by 9.628.438+9.6
- CritPtQwen3.8 Flash Next by 83.111.1+8
- GPQA DiamondQwen3.8 Flash Next by 5.686.792.3+5.6
FactsEven
- AA-Omniscience · Non-hallucinationNemotron 3 Ultra 550B A55B by 15.670.354.7+15.6
- AA-Omniscience · AccuracyQwen3.8 Flash Next by 1.922.624.5+1.9
MathNemotron 3 Ultra 550B A55B
- LiveBench · MathematicsNemotron 3 Ultra 550B A55B by 2.988.785.8+2.9
Following instructionsQwen3.8 Flash Next
- LiveBench · Instruction FollowingQwen3.8 Flash Next by 3.673.577.1+3.6
- IFBenchtie81.481.3tie
Other results21 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 [offline]Qwen3.8 Flash Next by 52.6355.6+52.6
- BrowseComp [offline]Qwen3.8 Flash Next by 454489+45
- GDPval-AA v2Qwen3.8 Flash Next by 413 rating points11481561+413 rating
- Toolathlon VerifiedQwen3.8 Flash Next by 39.234.373.5+39.2
- AA Agentic IndexQwen3.8 Flash Next by 34.721.756.4+34.7
- τ³-Bench BankingQwen3.8 Flash Next by 26.118.444.5+26.1
- Artificial Analysis Coding IndexQwen3.8 Flash Next by 23.849.373.1+23.8
- Terminal Bench 4 [offline]Qwen3.8 Flash Next by 20.7020.7+20.7
- livebench_data_analysisQwen3.8 Flash Next by 19.754.574.2+19.7
- AA IntelligenceQwen3.8 Flash Next by 16.922.939.8+16.9
- HLE, text-onlyQwen3.8 Flash Next by 10.429.239.6+10.4
- AA-OmniscienceNemotron 3 Ultra 550B A55B by 9.3-0.4-9.7+9.3
- tau2-bench telecomQwen3.8 Flash Next by 7.48390.4+7.4
- SciCode, subtaskQwen3.8 Flash Next by 4.64246.6+4.6
- livebench_languageQwen3.8 Flash Next by 3.870.874.6+3.8
- SciCode (subtask)Qwen3.8 Flash Next by 3.143.546.6+3.1
- WMT24++Qwen3.8 Flash Next by 2.583.786.2+2.5
- tau2-bench retailNemotron 3 Ultra 550B A55B by 2.173.771.6+2.1
- tau2-bench airlineQwen3.8 Flash Next by 179.880.8+1
- LMArena v3 proxytie34.335.2tie
- MMLU-ProXtie8383tie
Questions people ask
Which is better, Nemotron 3 Ultra 550B A55B or Qwen3.8 Flash Next?
Qwen3.8 Flash Next wins four of the seven areas where both have results: coding, agents, reasoning and following instructions. Nemotron 3 Ultra 550B A55B wins math. They are level on facts and long documents.
Which is better for coding?
Qwen3.8 Flash Next. It wins 8 of the 8 coding tests both models report; Nemotron 3 Ultra 550B A55B wins none.
Which is cheaper?
Nemotron 3 Ultra 550B A55B costs $0.60 per million input tokens and $2.70 per million output tokens; Qwen3.8 Flash Next costs $0.15 and $0.47. That makes Qwen3.8 Flash Next about 81% 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 22 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 21 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.