GPT-5.6 Luna vs Qwen3.5 35B A3B
Wins 5 of 7 areas
Coding · Agents · Reasoning · Math · Long documents
Wins 0 of 7 areas
—
GPT-5.6 Luna is the stronger all-rounder.
Scores updated · 35 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 software50GPT-5.6 Luna5 of 5 tests
- AgentsCarrying out multi-step tasks on its own50GPT-5.6 Luna5 of 5 tests
- ReasoningHard problems that need careful thinking30GPT-5.6 Luna3 of 3 tests
- MathCompetition and research-level math20GPT-5.6 Luna2 of 2 tests
- Long documentsFinding answers in very long texts10GPT-5.6 Luna1 of 1 test
- FactsGetting facts right instead of making them up11Even1 each
- Images and chartsUnderstanding pictures, charts and video22Even2 each
Long documents 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.
GPT-5.6 Luna costs 38% less for the same work.
The biggest differences
The three tests each model wins by the widest margin. Scores are out of 100.
Where GPT-5.6 Luna pulls ahead
- Real work tasks from 44 professionsGDPVal+42.6points ahead
- Command-line tasks in a real terminalTerminal-Bench 2.1+40.1points ahead
- Customer-service tasks in a simulated bankτ-Bench V3 · Banking+26.2points ahead
Where Qwen3.5 35B A3B pulls ahead
- Questions that truly need the image to answerMMStar+9.7points ahead
- Math problems shown in pictures and chartsMathVista+9.3points ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+7.2points ahead
Every test, side by side
All 35 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingGPT-5.6 Luna
- Terminal-Bench 2.1GPT-5.6 Luna by 40.180.940.8+40.1
- LMArena · WebDevGPT-5.6 Luna by 269 rating points15191250+269 rating
- SWE-bench VerifiedGPT-5.6 Luna by 23.89369.2+23.8
- SWE-bench ProGPT-5.6 Luna by 18.162.744.6+18.1
- SciCodeGPT-5.6 Luna by 15.953.637.7+15.9
AgentsGPT-5.6 Luna
- GDPValGPT-5.6 Luna by 42.6485.4+42.6
- τ-Bench V3 · BankingGPT-5.6 Luna by 26.231.14.9+26.2
- BrowseCompGPT-5.6 Luna by 22.383.361+22.3
- AA IT-Bench SREGPT-5.6 Luna by 18.840.321.5+18.8
- OSWorld-VerifiedGPT-5.6 Luna by 18.172.654.5+18.1
ReasoningGPT-5.6 Luna
- CritPtGPT-5.6 Luna by 19.720.60.9+19.7
- Humanity's Last ExamGPT-5.6 Luna by 18.539.521+18.5
- GPQA DiamondGPT-5.6 Luna by 6.691.184.5+6.6
FactsEven
- AA-Omniscience · AccuracyGPT-5.6 Luna by 22.642.720.1+22.6
- AA-Omniscience · Non-hallucinationQwen3.5 35B A3B by 7.27.414.6+7.2
Images and chartsEven
- MMStarQwen3.5 35B A3B by 9.772.281.9+9.7
- MathVistaQwen3.5 35B A3B by 9.376.986.2+9.3
- MMMU-ProGPT-5.6 Luna by 5.978.672.7+5.9
- CharXiv (RQ)GPT-5.6 Luna by 5.282.777.5+5.2
MathGPT-5.6 Luna
- HMMT Feb. 2026GPT-5.6 Luna by 16.798.581.8+16.7
- AIME 2026GPT-5.6 Luna by 4.397.693.3+4.3
Other results13 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.
- AA-OmniscienceGPT-5.6 Luna by 37.8-10.3-48.1+37.8
- Artificial Analysis Coding IndexGPT-5.6 Luna by 34.471.437+34.4
- AA Agentic IndexGPT-5.6 Luna by 30.942.711.8+30.9
- AA IntelligenceGPT-5.6 Luna by 1837.319.3+18
- LVBenchQwen3.5 35B A3B by 12.558.971.4+12.5
- RealWorldQAQwen3.5 35B A3B by 11.772.484.1+11.7
- RefSpatial-BenchQwen3.5 35B A3B by 1152.563.5+11
- HallusionBenchQwen3.5 35B A3B by 8.95967.9+8.9
- ERQAQwen3.5 35B A3B by 8.356.564.8+8.3
- EmbSpatial-BenchQwen3.5 35B A3B by 7.675.583.1+7.6
- MV-BenchQwen3.5 35B A3B by 6.668.274.8+6.6
- SimpleVQAQwen3.5 35B A3B by 5.153.258.3+5.1
- VideoMMMUtie81.180.4tie
Questions people ask
Which is better, GPT-5.6 Luna or Qwen3.5 35B A3B?
GPT-5.6 Luna wins five of the seven areas where both have results: coding, agents, reasoning, math and long documents. Qwen3.5 35B A3B wins none. They are level on facts and images and charts.
Which is better for coding?
GPT-5.6 Luna. It wins 5 of the 5 coding tests both models report; Qwen3.5 35B A3B wins none.
Which is cheaper?
GPT-5.6 Luna costs $0.20 per million input tokens and $1.20 per million output tokens; Qwen3.5 35B A3B costs $0.25 and $2.00. That makes GPT-5.6 Luna about 38% cheaper for the same work.
How do you compare the two?
We use the 35 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 13 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.