The two are evenly matched.LFM2.5-1.2B-Thinking is better at following instructions; Qwen3 1.7B at facts.
Scores updated · 20 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.
- Following instructionsDoing exactly what it is asked10LFM2.5-1.2B-Thinking1 of 1 test
- FactsGetting facts right instead of making them up01Qwen3 1.7B1 of 2 tests · 1 tie
- CodingWriting and fixing software00Even0 each · 2 ties
- AgentsCarrying out multi-step tasks on its own00Even0 each · 1 tie
- ReasoningHard problems that need careful thinking11Even1 each · 1 tie
- Long documentsFinding answers in very long texts00Even0 each · 1 tie
Agents, long documents and following instructions rest on a single test each.
The biggest differences
The tests each model wins by the widest margin, up to three each. Scores are out of 100.
Where LFM2.5-1.2B-Thinking pulls ahead
- Follows unfamiliar, precisely checkable instructionsIFBench+14.9points ahead
- Very hard expert questions across many subjectsHumanity's Last Exam+1.6points ahead
Where Qwen3 1.7B pulls ahead
- Graduate-level biology, physics and chemistry questionsGPQA Diamond+1.7points ahead
- Answers hard knowledge questions correctlyAA-Omniscience · Accuracy+1.5points ahead
Every test, side by side
All 20 tests both models report. The winning score is in its model's colour; marks a score checked independently.
ReasoningEven
- GPQA DiamondQwen3 1.7B by 1.733.935.6+1.7
- Humanity's Last ExamLFM2.5-1.2B-Thinking by 1.66.24.6+1.6
- CritPttie00tie
FactsQwen3 1.7B
- AA-Omniscience · AccuracyQwen3 1.7B by 1.57.48.9+1.5
- AA-Omniscience · Non-hallucinationtie5.65.2tie
Following instructionsLFM2.5-1.2B-Thinking
- IFBenchLFM2.5-1.2B-Thinking by 14.941.826.9+14.9
Other results10 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.
- GSM8KLFM2.5-1.2B-Thinking by 34.285.651.4+34.2
- IFEvalLFM2.5-1.2B-Thinking by 14.488.474+14.4
- MMLU-ProQwen3 1.7B by 7.349.657+7.3
- τ²-Bench Telecom (AA run)Qwen3 1.7B by 6.419.626+6.4
- MATH-500 (EM)LFM2.5-1.2B-Thinking by 6.18881.9+6.1
- AIME25 no toolsQwen3 1.7B by 4.631.736.3+4.6
- AA-OmniscienceQwen3 1.7B by 2.4-80-77.5+2.4
- AA Agentic IndexQwen3 1.7B by 2.26.58.7+2.2
- BFCL v3LFM2.5-1.2B-Thinking by 1.65755.4+1.6
- AA Intelligencetie5.25.2tie
Questions people ask
Which is better, LFM2.5-1.2B-Thinking or Qwen3 1.7B?
LFM2.5-1.2B-Thinking and Qwen3 1.7B each win one of the six areas where both have results. LFM2.5-1.2B-Thinking wins following instructions; Qwen3 1.7B wins facts. They are level on coding, agents, reasoning and long documents.
Which is better for coding?
Neither. All 2 coding tests both models report are ties.
How do you compare the two?
We use the 20 benchmark tests both models have published scores on. The verdict counts the 10 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 10 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.