Qwen3.5 2B
- Input
- —
- Output
- —
- Speed
- ~93tok/sest.
- Context
- 262Ktokens
Long-tail import from Artificial Analysis (2026-09-24); AA variant: Qwen3.5 2B (Non-reasoning). Price = AA median host price.
Rankings
Writing & chat
Verified#131 / 162rank by quality21.9±9.5quality / 1001/5 benchmarks measuredVision
Verified#87 / 98rank by quality19.7±8.2quality / 1001/6 benchmarks measuredHard reasoning
Verified#182 / 218rank by quality15.9±5.9quality / 1002/4 benchmarks measuredAgents
Verified#84 / 187rank by quality36.8±10.0quality / 1001/7 benchmarks measured
Self-hostingEstimated
How we estimate →What it would cost to run these open weights yourself on rented GPUs. No API sells this model, so this is its price on the chart.
- Hardware
- 1× B200 180GB
- BF16 weights
- Throughput
- ~22,882 tok/s
- many requests batched
- Price
- $0.030–0.27 /M tok
- busy → light use
- On your own machine
- RTX 5090 32GB (int4, 300+ tok/s single-stream)
- single consumer GPU or Mac
Assumptions (5)
- BF16 weights (2B params) + 25% KV-cache headroom ≈ 5 GB
- 1× B200 180GB at $5.98–$14.24/GPU-hour on-demand (2026-09-24)
- ~22,882 output tok/s aggregate at batch 256 (bandwidth-bound)
- Blended 3:1 input:output; prefill ~265,781 tok/s
- Low = 75% utilization at the low GPU price; high = 20% at the high price
Benchmark scores
~ italic, dashed = no published score yet, estimated from related benchmarks.
Writing & chat
IFBenchVerified
#118 of 13929.1%
- #1MiniMax-M382.9%
- #2Qwen3.8 2.4T A95B82.8%
- #118Ministral 3 8B29.1%
- #118Qwen3.5 2B29.1%
- #120Hermes 4 - Llama-3.1 70B29.0%
artificialanalysis.ai · 2026-09-24
MMMLUEstimated
~last of 5 (estimated)below measured range (<81.3%)
No published score — estimated from related benchmarks
LMArena Text, Non-English (Elo)Estimated
~#48 of 57 (estimated)~1325 Elo
- #1Gemini 4 ArgonFew votes1512
- #2Claude Opus 5.51497
- #47INTELLECT-3Few votes1327
- ~#48Qwen3.5 2BEstimated1325
- #48Llama 4 Maverick1311
No published score — estimated from related benchmarks
LMArena Text (Elo)Estimated
~#49 of 57 (estimated)~1348 Elo
- #1Gemini 4 Argon1525
- #2Claude Opus 5.51507
- #48Llama 3.1 Nemotron Ultra 253B v1Few votes1348
- ~#49Qwen3.5 2BEstimated1348
- #49Olmo 3.1 32B Instruct1330
No published score — estimated from related benchmarks
EQ-Bench Creative Writing v3 (Elo)Estimated
~#23 of 26 (estimated)~1373 Elo
- #1GPT-6 Astra2173
- #2Claude Fable 5.12162
- #22Mistral Large 31412
- ~#23Qwen3.5 2BEstimated1373
- #23Gemma 4 31B1368
No published score — estimated from related benchmarks
Vision
CharXiv ReasoningEstimated
~last of 12 (estimated)below measured range (<74.5%)
No published score — estimated from related benchmarks
OmniDocBench v1.5Estimated
~last of 8 (estimated)below measured range (<75.8%)
No published score — estimated from related benchmarks
GDP.pdf (AA)Estimated
~#46 of 48 (estimated)~0.2%
- #1Muse Wonder Full32.6%
- #2GPT-6 Astra31.0%
- #45Ministral 3 8B0.8%
- ~#46Qwen3.5 2BEstimated0.2%
- #46Ministral 3 3B0.0%
No published score — estimated from related benchmarks
ChartographyEstimated
~last of 25 (estimated)below measured range (<9.0%)
No published score — estimated from related benchmarks
Hard reasoning
GPQA DiamondVerified
#162 of 19243.8%
- #1GPT-6 Astra96.1%
- #2Gemini 3.8 Flash95.3%
- #161Phi-4-mini-flash-reasoning45.1%
- #162Qwen3.5 2B43.8%
- #163Llama 3.2 Instruct 90B (Vision)43.2%
artificialanalysis.ai · 2026-09-24
Humanity's Last ExamVerified
#162 of 1985.0%
- #1Claude Fable 5.274.3%
- #2Claude Opus 5.561.4%
- #162Phi-4 Multimodal Instruct5.0%
- #162Qwen3.5 2B5.0%
- #169Llama 4 Maverick4.9%
artificialanalysis.ai · 2026-09-24
AIME (latest)Estimated
~#27 of 34 (estimated)~32.1%
- #1GLM-5.3 NVFP499.5%
- #2Beam 1.0 501B A23B97.7%
- #26Phi-4-mini-flash-reasoning33.6%
- ~#27Qwen3.5 2BEstimated32.1%
- #27Ministral 3 8B31.7%
No published score — estimated from related benchmarks
AA Intelligence IndexEstimated
~#96 of 103 (estimated)~7.3
- #1Gemini 4 Carbon59.4
- #2Claude Opus 5.557.6
- #95Devstral Small 27.5
- ~#96Qwen3.5 2BEstimated7.3
- #96Llama 4 Scout6.5
No published score — estimated from related benchmarks
Agents
τ²-bench (Telecom)Verified
#28 of 11981.6%
- #1Step 3.7 Flash98.5%
- #2DeepSeek-V4-Pro-0813-nvfp4-DSpark98.3%
- #28Nex-N2-Pro81.6%
- #28Qwen3.5 2B81.6%
- #30MiniCPM5-1B81.0%
artificialanalysis.ai · 2026-09-24
OSWorld-VerifiedEstimated
~#12 of 15 (estimated)~60.2%
- #1Kimi K384.8%
- #2Qwen3.8 27B84.3%
- #11Muse Glimmer65.9%
- ~#12Qwen3.5 2BEstimated60.2%
- #12Nemotron 3.5 Super VL NVFP452.3%
No published score — estimated from related benchmarks
OSWorld 2.0Estimated
~#3 of 5 (estimated)~53.8%
No published score — estimated from related benchmarks
BrowseCompEstimated
~#19 of 22 (estimated)~61.2%
- #1Atria Dawn Preview92.5%
- #2GPT-6 Astra91.5%
- #18Nemotron 3.5 Bolt 74B-A9B70.0%
- ~#19Qwen3.5 2BEstimated61.2%
- #19Mistral Medium 3.548.6%
No published score — estimated from related benchmarks
GDPval-AA (Elo)Estimated
~#47 of 92 (estimated)~888 Elo
- #1Claude Opus 5.51866
- #2Muse Wonder Full1857
- #46GPT-5.4 nano937
- ~#47Qwen3.5 2BEstimated888
- #47Qwen3.6 35B A3B879
No published score — estimated from related benchmarks
AutomationBench-AAEstimated
~#61 of 85 (estimated)~3.0%
- #1Claude Fable 5.276.4%
- #2Muse Wonder Full71.3%
- #60Nemotron 3 Ultra3.0%
- ~#61Qwen3.5 2BEstimated3.0%
- #61Mercury 2.52.4%
No published score — estimated from related benchmarks
τ-Bench Banking (AA)Estimated
~#80 of 91 (estimated)~5.4%
- #1Muse Spark 1.350.5%
- #2GLM-5.350.3%
- #78HyperNova 60B 26055.6%
- ~#80Qwen3.5 2BEstimated5.4%
- #80Qwen3 Coder Next5.4%
No published score — estimated from related benchmarks