Qwen2.5 Instruct 72B vs Qwen3.5 122B A10B
Wins 0 of 5 areas
—
Wins 4 of 5 areas
Coding · Reasoning · Long documents · Following instructions
Qwen3.5 122B A10B is the stronger all-rounder.Qwen2.5 Instruct 72B is cheaper.
Scores updated · 20 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 software03Qwen3.5 122B A10B3 of 3 tests
- ReasoningHard problems that need careful thinking02Qwen3.5 122B A10B2 of 3 tests · 1 tie
- Long documentsFinding answers in very long texts02Qwen3.5 122B A10B2 of 2 tests
- Following instructionsDoing exactly what it is asked01Qwen3.5 122B A10B1 of 1 test
- FactsGetting facts right instead of making them up11Even1 each
Following instructions rests on a single test.
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.
Qwen2.5 Instruct 72B costs 73% 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 Qwen3.5 122B A10B pulls ahead
- Reasons across sets of long documentsAA-LCR+56.3points ahead
- Fixes real GitHub issues in Python projectsSWE-bench Verified+48.2points ahead
- Follows unfamiliar, precisely checkable instructionsIFBench+38.8points ahead
Where Qwen2.5 Instruct 72B pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+1.6points 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.
CodingQwen3.5 122B A10B
- SWE-bench VerifiedQwen3.5 122B A10B by 48.223.872+48.2
- Terminal-Bench HardQwen3.5 122B A10B by 26.64.531.1+26.6
- SciCodeQwen3.5 122B A10B by 1326.739.7+13
ReasoningQwen3.5 122B A10B
- GPQA DiamondQwen3.5 122B A10B by 36.649.185.7+36.6
- Humanity's Last ExamQwen3.5 122B A10B by 21.63.625.2+21.6
- CritPttie00.6tie
FactsEven
- AA-Omniscience · AccuracyQwen3.5 122B A10B by 6.917.524.4+6.9
- AA-Omniscience · Non-hallucinationQwen2.5 Instruct 72B by 1.614.512.9+1.6
Long documentsQwen3.5 122B A10B
- AA-LCRQwen3.5 122B A10B by 56.32076.3+56.3
- LongBench v2Qwen3.5 122B A10B by 20.839.460.2+20.8
Following instructionsQwen3.5 122B A10B
- IFBenchQwen3.5 122B A10B by 38.836.975.7+38.8
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.
- τ²-Bench Telecom (AA run)Qwen3.5 122B A10B by 59.134.593.6+59.1
- Artificial Analysis Coding IndexQwen3.5 122B A10B by 33.811.945.7+33.8
- MMLU-ProQwen3.5 122B A10B by 15.671.186.7+15.6
- MMMLUQwen3.5 122B A10B by 11.974.886.7+11.9
- AA-OmniscienceQwen3.5 122B A10B by 11.6-53.1-41.5+11.6
- IFEvalQwen3.5 122B A10B by 9.384.193.4+9.3
- AA IntelligenceQwen3.5 122B A10B by 7.97.715.6+7.9
- MMLU-ReduxQwen3.5 122B A10B by 7.286.894+7.2
- C-EvalQwen3.5 122B A10B by 2.789.291.9+2.7
Questions people ask
Which is better, Qwen2.5 Instruct 72B or Qwen3.5 122B A10B?
Qwen3.5 122B A10B wins four of the five areas where both have results: coding, reasoning, long documents and following instructions. Qwen2.5 Instruct 72B wins none, but costs 73% less. They are level on facts.
Which is better for coding?
Qwen3.5 122B A10B. It wins 3 of the 3 coding tests both models report; Qwen2.5 Instruct 72B wins none.
Which is cheaper?
Qwen2.5 Instruct 72B costs $0.47 per million input tokens and $0.49 per million output tokens; Qwen3.5 122B A10B costs $0.40 and $3.20. That makes Qwen2.5 Instruct 72B about 73% cheaper for the same work.
How do you compare the two?
We use the 20 benchmark tests both models have published scores on. The verdict counts the 11 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.