Claude Opus 4.8 vs Qwen3.5 35B A3B
Wins 7 of 8 areas
Coding · Agents · Reasoning · Facts · Images and charts · Math · Long documents
Wins 1 of 8 areas
Following instructions
Claude Opus 4.8 is the stronger all-rounder.Qwen3.5 35B A3B is cheaper and better at following instructions.
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.
- CodingWriting and fixing software70Claude Opus 4.87 of 7 tests
- AgentsCarrying out multi-step tasks on its own50Claude Opus 4.85 of 5 tests
- ReasoningHard problems that need careful thinking30Claude Opus 4.83 of 3 tests
- MathCompetition and research-level math30Claude Opus 4.83 of 3 tests
- FactsGetting facts right instead of making them up20Claude Opus 4.82 of 2 tests
- Images and chartsUnderstanding pictures, charts and video10Claude Opus 4.81 of 1 test
- Long documentsFinding answers in very long texts10Claude Opus 4.81 of 1 test
- Following instructionsDoing exactly what it is asked01Qwen3.5 35B A3B1 of 1 test
Images and charts, 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 93% 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 Claude Opus 4.8 pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+46.1points ahead
- Command-line tasks in a real terminalTerminal-Bench 2.1+43.8points ahead
- Real work tasks from 44 professionsGDPVal+42.2points ahead
Where Qwen3.5 35B A3B pulls ahead
- Follows unfamiliar, precisely checkable instructionsIFBench+10.3points 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.
CodingClaude Opus 4.8
- Terminal-Bench 2.1Claude Opus 4.8 by 43.884.640.8+43.8
- Terminal-Bench HardClaude Opus 4.8 by 31.858.326.5+31.8
- LMArena · WebDevClaude Opus 4.8 by 306 rating points15561250+306 rating
- SWE-bench ProClaude Opus 4.8 by 24.669.244.6+24.6
- SWE-bench MultilingualClaude Opus 4.8 by 24.184.460.3+24.1
- SWE-bench VerifiedClaude Opus 4.8 by 19.488.669.2+19.4
- SciCodeClaude Opus 4.8 by 16.754.437.7+16.7
AgentsClaude Opus 4.8
- GDPValClaude Opus 4.8 by 42.247.85.6+42.2
- τ-Bench V3 · BankingClaude Opus 4.8 by 29.334.24.9+29.3
- OSWorld-VerifiedClaude Opus 4.8 by 28.983.454.5+28.9
- BrowseCompClaude Opus 4.8 by 23.384.361+23.3
- MCP AtlasClaude Opus 4.8 by 19.882.262.4+19.8
ReasoningClaude Opus 4.8
- Humanity's Last ExamClaude Opus 4.8 by 27.748.721+27.7
- CritPtClaude Opus 4.8 by 2020.90.9+20
- GPQA DiamondClaude Opus 4.8 by 7.59284.5+7.5
FactsClaude Opus 4.8
- AA-Omniscience · Non-hallucinationClaude Opus 4.8 by 46.160.714.6+46.1
- AA-Omniscience · AccuracyClaude Opus 4.8 by 28.748.820.1+28.7
Images and chartsClaude Opus 4.8
- CharXiv (RQ)Claude Opus 4.8 by 12.489.977.5+12.4
MathClaude Opus 4.8
- HMMT Feb. 2026Claude Opus 4.8 by 13.795.581.8+13.7
- AIME 2026Claude Opus 4.8 by 6.710093.3+6.7
- IMOAnswerBenchClaude Opus 4.8 by 6.783.576.8+6.7
Following instructionsQwen3.5 35B A3B
- IFBenchQwen3.5 35B A3B by 10.362.272.5+10.3
Other results20 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-OmniscienceClaude Opus 4.8 by 76.928.8-48.1+76.9
- NL2RepoClaude Opus 4.8 by 49.269.720.5+49.2
- Artificial Analysis Coding IndexClaude Opus 4.8 by 37.374.337+37.3
- Terminal-Bench 2.0Claude Opus 4.8 by 34.174.640.5+34.1
- Tool-DecathlonClaude Opus 4.8 by 31.259.928.7+31.2
- AA Agentic IndexClaude Opus 4.8 by 30.842.611.8+30.8
- AA IntelligenceClaude Opus 4.8 by 22.541.819.3+22.5
- ScreenSpot-Pro (No tools)Claude Opus 4.8 by 19.387.968.6+19.3
- AgentWorldBench - SWEClaude Opus 4.8 by 16.564.147.6+16.5
- WideSearchClaude Opus 4.8 by 15.872.957.1+15.8
- AgentWorldBench - TerminalClaude Opus 4.8 by 13.159.246.1+13.1
- AgentWorldBench - OSClaude Opus 4.8 by 10.366.656.3+10.3
- AgentWorldBench - SearchClaude Opus 4.8 by 9.135.126+9.1
- AgentWorldBench - OverallClaude Opus 4.8 by 8.956.647.7+8.9
- AgentWorldBench - AndroidClaude Opus 4.8 by 8.361.553.2+8.3
- IncludeClaude Opus 4.8 by 7.987.679.7+7.9
- AgentWorldBench - WebClaude Opus 4.8 by 7.654.747.1+7.6
- HMMT Nov. 2025Claude Opus 4.8 by 7.396.589.2+7.3
- τ²-Bench Telecom (AA run)Claude Opus 4.8 by 5.294.489.2+5.2
- AgentWorldBench - MCPQwen3.5 35B A3B by 354.957.9+3
Questions people ask
Which is better, Claude Opus 4.8 or Qwen3.5 35B A3B?
Claude Opus 4.8 wins seven of the eight areas we test: coding, agents, reasoning, facts, images and charts, math and long documents. Qwen3.5 35B A3B wins following instructions, and costs 93% less.
Which is better for coding?
Claude Opus 4.8. It wins 7 of the 7 coding tests both models report; Qwen3.5 35B A3B wins none.
Which is cheaper?
Claude Opus 4.8 costs $5.00 per million input tokens and $25.00 per million output tokens; Qwen3.5 35B A3B costs $0.25 and $2.00. That makes Qwen3.5 35B A3B about 93% 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 23 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 20 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.