GPT-5.5 vs Qwen3.5 35B A3B
Wins 7 of 8 areas
Coding · Agents · Reasoning · Facts · Math · Long documents · Following instructions
Wins 1 of 8 areas
Images and charts
GPT-5.5 is the stronger all-rounder.Qwen3.5 35B A3B is cheaper and better at images and charts.
Scores updated · 43 tests both models report · How we compare
Where each one wins
Tests won in each of the eight areas we test. Each piece is one test, so a longer bar means more evidence; grey means the two scored within a point of each other.
- AgentsCarrying out multi-step tasks on its own60GPT-5.56 of 6 tests
- CodingWriting and fixing software50GPT-5.55 of 5 tests
- ReasoningHard problems that need careful thinking30GPT-5.53 of 3 tests
- MathCompetition and research-level math20GPT-5.52 of 2 tests
- FactsGetting facts right instead of making them up21GPT-5.52 of 3 tests
- Long documentsFinding answers in very long texts10GPT-5.51 of 1 test
- Following instructionsDoing exactly what it is asked10GPT-5.51 of 1 test
- Images and chartsUnderstanding pictures, charts and video12Qwen3.5 35B A3B2 of 3 tests
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.
Qwen3.5 35B A3B costs 94% 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-5.5 pulls ahead
- Command-line tasks in a real terminalTerminal-Bench 2.1+43.5points ahead
- Answers hard knowledge questions correctlyAA-Omniscience · Accuracy+37.9points ahead
- Real work tasks from 44 professionsGDPVal+37.1points ahead
Where Qwen3.5 35B A3B pulls ahead
- Reads and reasons about text in imagesOCRBenchv2+4.2points ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+3.6points ahead
- Math problems shown in pictures and chartsMathVista+2points 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.
CodingGPT-5.5
- Terminal-Bench 2.1GPT-5.5 by 43.584.340.8+43.5
- Terminal-Bench HardGPT-5.5 by 34.160.626.5+34.1
- LMArena · WebDevGPT-5.5 by 262 rating points15121250+262 rating
- SciCodeGPT-5.5 by 18.155.837.7+18.1
- SWE-bench ProGPT-5.5 by 1458.644.6+14
AgentsGPT-5.5
- GDPValGPT-5.5 by 37.142.75.6+37.1
- τ-Bench V3 · BankingGPT-5.5 by 34.1394.9+34.1
- AA IT-Bench SREGPT-5.5 by 24.345.821.5+24.3
- OSWorld-VerifiedGPT-5.5 by 24.278.754.5+24.2
- BrowseCompGPT-5.5 by 23.484.461+23.4
- MCP AtlasGPT-5.5 by 12.975.362.4+12.9
ReasoningGPT-5.5
- CritPtGPT-5.5 by 26.227.10.9+26.2
- Humanity's Last ExamGPT-5.5 by 24.845.821+24.8
- GPQA DiamondGPT-5.5 by 993.584.5+9
FactsGPT-5.5
- AA-Omniscience · AccuracyGPT-5.5 by 37.95820.1+37.9
- AA-Omniscience · Non-hallucinationQwen3.5 35B A3B by 3.61114.6+3.6
- Vectara HHEM hallucination ratelower is betterGPT-5.5 by 1.29.310.5+1.2
Images and chartsQwen3.5 35B A3B
- MMMU-ProGPT-5.5 by 7.279.972.7+7.2
- OCRBenchv2Qwen3.5 35B A3B by 4.261.165.3+4.2
- MathVistaQwen3.5 35B A3B by 284.286.2+2
MathGPT-5.5
- HMMT Feb. 2026GPT-5.5 by 16.798.581.8+16.7
- AIME 2026GPT-5.5 by 6.710093.3+6.7
Other results19 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-5.5 by 68.620.5-48.1+68.6
- Terminal-Bench 2.0GPT-5.5 by 42.282.740.5+42.2
- Artificial Analysis Coding IndexGPT-5.5 by 37.974.937+37.9
- NL2RepoGPT-5.5 by 30.250.720.5+30.2
- Tool-DecathlonGPT-5.5 by 26.955.628.7+26.9
- AA Agentic IndexGPT-5.5 by 25.537.311.8+25.5
- AA IntelligenceGPT-5.5 by 19.138.419.3+19.1
- BabyVisionGPT-5.5 by 17.555.938.4+17.5
- SuperGPQAGPT-5.5 by 9.372.763.4+9.3
- DynaMathQwen3.5 35B A3B by 9.175.985+9.1
- HMMT Nov. 2025GPT-5.5 by 7.396.589.2+7.3
- τ²-Bench Telecom (AA run)GPT-5.5 by 4.793.989.2+4.7
- vectara_avg_summary_lengthGPT-5.5 by 34.7 rating points129.694.9+34.7 rating
- RealWorldQAQwen3.5 35B A3B by 1.982.284.1+1.9
- EmbSpatial-BenchQwen3.5 35B A3B by 1.281.983.1+1.2
- vectara_factual_consistencyGPT-5.5 by 1.290.789.5+1.2
- ERQAtie64.564.8tie
- SimpleVQAtie58.658.3tie
- vectara_answer_ratetie10099.8tie
Questions people ask
Which is better, GPT-5.5 or Qwen3.5 35B A3B?
GPT-5.5 wins seven of the eight areas we test: coding, agents, reasoning, facts, math, long documents and following instructions. Qwen3.5 35B A3B wins images and charts, and costs 94% less.
Which is better for coding?
GPT-5.5. It wins 5 of the 5 coding tests both models report; Qwen3.5 35B A3B wins none.
Which is cheaper?
GPT-5.5 costs $5.00 per million input tokens and $30.00 per million output tokens; Qwen3.5 35B A3B costs $0.25 and $2.00. That makes Qwen3.5 35B A3B about 94% 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 24 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 19 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.