Claude Opus 4.8 vs Qwen3.8 Flash Next
Wins 4 of 8 areas
Coding · Reasoning · Facts · Math
Wins 3 of 8 areas
Agents · Long documents · Following instructions
Claude Opus 4.8 wins more areas, narrowly.Qwen3.8 Flash Next is cheaper and better at agents.
Both rank among the ten best models we track in coding.
Scores updated · 39 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.
- ReasoningHard problems that need careful thinking30Claude Opus 4.83 of 4 tests · 1 tie
- CodingWriting and fixing software53Claude Opus 4.85 of 8 tests
- FactsGetting facts right instead of making them up20Claude Opus 4.82 of 2 tests
- MathCompetition and research-level math10Claude Opus 4.81 of 1 test
- AgentsCarrying out multi-step tasks on its own14Qwen3.8 Flash Next4 of 5 tests
- Following instructionsDoing exactly what it is asked02Qwen3.8 Flash Next2 of 2 tests
- Long documentsFinding answers in very long texts01Qwen3.8 Flash Next1 of 1 test
- Images and chartsUnderstanding pictures, charts and video00Even0 each · 1 tie
Images and charts, math and long documents 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.
Qwen3.8 Flash Next costs 98% less for the same work.
The biggest differences
The three tests each model wins by the widest margin. Scores are out of 100.
Where Claude Opus 4.8 pulls ahead
- Completes tasks by operating a computer desktopOSWorld-Verified+27.6points ahead
- Answers hard knowledge questions correctlyAA-Omniscience · Accuracy+24.3points ahead
- Very hard expert questions across many subjectsHumanity's Last Exam+10.7points ahead
Where Qwen3.8 Flash Next pulls ahead
- Follows unfamiliar, precisely checkable instructionsIFBench+19.1points ahead
- Customer-service tasks in a simulated bankτ-Bench V3 · Banking+11.2points ahead
- Fixes real GitHub issues, working as an agentLiveBench · Agentic Coding+11.1points ahead
Every test, side by side
All 39 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingClaude Opus 4.8
- LiveBench · Agentic CodingQwen3.8 Flash Next by 11.150.561.6+11.1
- SWE-bench VerifiedClaude Opus 4.8 by 9.988.678.7+9.9
- LiveBench · CodingClaude Opus 4.8 by 9.281.872.5+9.2
- LMArena · WebDevQwen3.8 Flash Next by 72 rating points15561628+72 rating
- SWE-bench ProClaude Opus 4.8 by 6.769.262.5+6.7
- SciCodeClaude Opus 4.8 by 3.854.450.6+3.8
- SWE-bench MultilingualClaude Opus 4.8 by 3.484.481+3.4
- Terminal-Bench 2.1Qwen3.8 Flash Next by 1.584.686.1+1.5
AgentsQwen3.8 Flash Next
- OSWorld-VerifiedClaude Opus 4.8 by 27.683.455.8+27.6
- τ-Bench V3 · BankingQwen3.8 Flash Next by 11.234.245.4+11.2
- GDPValQwen3.8 Flash Next by 8.747.756.4+8.7
- BrowseCompQwen3.8 Flash Next by 4.784.389+4.7
- Terminal-Bench 4.0Qwen3.8 Flash Next by 3.621.725.3+3.6
ReasoningClaude Opus 4.8
- Humanity's Last ExamClaude Opus 4.8 by 10.748.738+10.7
- CritPtClaude Opus 4.8 by 9.820.911.1+9.8
- LiveBench · ReasoningClaude Opus 4.8 by 1.889.287.4+1.8
- GPQA Diamondtie9292.3tie
FactsClaude Opus 4.8
- AA-Omniscience · AccuracyClaude Opus 4.8 by 24.348.824.5+24.3
- AA-Omniscience · Non-hallucinationClaude Opus 4.8 by 660.754.7+6
Following instructionsQwen3.8 Flash Next
- IFBenchQwen3.8 Flash Next by 19.162.281.3+19.1
- LiveBench · Instruction FollowingQwen3.8 Flash Next by 5.17277.1+5.1
Other results15 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 38.528.8-9.7+38.5
- OSWorld 2.0Claude Opus 4.8 by 36.355.719.4+36.3
- Agents' Last ExamQwen3.8 Flash Next by 25.525.751.2+25.5
- NL2RepoClaude Opus 4.8 by 21.669.748.1+21.6
- AA Agentic IndexQwen3.8 Flash Next by 13.842.656.4+13.8
- ToolathlonQwen3.8 Flash Next by 13.659.973.5+13.6
- HLE, full setClaude Opus 4.8 by 1149.838.8+11
- livebench_data_analysisQwen3.8 Flash Next by 8.26674.2+8.2
- JobBenchQwen3.8 Flash Next by 7.348.455.7+7.3
- Toolathlon VerifiedClaude Opus 4.8 by 6.479.973.5+6.4
- livebench_languageClaude Opus 4.8 by 5.179.774.6+5.1
- GDPval-AA v2Claude Opus 4.8 by 32 rating points15931561+32 rating
- AA IntelligenceClaude Opus 4.8 by 241.839.8+2
- Artificial Analysis Coding IndexClaude Opus 4.8 by 1.274.373.1+1.2
- DeepSWE 1.1tie5958.7tie
Questions people ask
Which is better, Claude Opus 4.8 or Qwen3.8 Flash Next?
Claude Opus 4.8 wins four of the eight areas we test: coding, reasoning, facts and math. Qwen3.8 Flash Next wins agents, long documents and following instructions, and costs 98% less. They are level on images and charts.
Which is better for coding?
Claude Opus 4.8. It wins 5 of the 8 coding tests both models report; Qwen3.8 Flash Next wins 3.
Which is cheaper?
Claude Opus 4.8 costs $5.00 per million input tokens and $25.00 per million output tokens; Qwen3.8 Flash Next costs $0.15 and $0.47. That makes Qwen3.8 Flash Next about 98% cheaper for the same work.
How do you compare the two?
We use the 39 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 15 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.