Intern-S2-Preview (35B-A3B)
- Input
- —
- Output
- —
- Speed
- ~110tok/sest.
- Context
- 262Ktokens
35B-A3B scientific multimodal model continued-pretrained from Qwen3.5. No API price found. Lab-reported scores from the HF model card results table (image; OpenCompass / VLMEvalKit, thinking mode).
Rankings
Coding
Lab-reported#101 / 137rank by quality33.2±8.6quality / 1002/7 benchmarks measuredWriting & chat
Lab-reported#39 / 162rank by quality66.0±9.7quality / 1001/5 benchmarks measuredVision
Lab-reported#33 / 98rank by quality64.2±8.7quality / 1001/6 benchmarks measuredHard reasoning
Lab-reported#92 / 218rank by quality50.0±8.7quality / 1001/4 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× H200 141GB
- FP8 weights
- Throughput
- ~8,390 tok/s
- many requests batched
- Price
- $0.046–0.38 /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)
- FP8 weights (35B params, 3B active per token) + 25% KV-cache headroom ≈ 44 GB
- 1× H200 141GB at $3.59–$7.91/GPU-hour on-demand (2026-09-24)
- ~8,390 output tok/s aggregate at batch 256 (bandwidth-bound); MoE compute scales with active params
- Blended 3:1 input:output; prefill ~155,846 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.
Coding
SWE-bench VerifiedLab-reported
#17 of 2064.0%
- #1Claude Opus 596.0%
- #2Claude Sonnet 585.2%
- #16Nemotron 3 Ultra70.7%
- #17Intern-S2-Preview (35B-A3B)64.0%
- #18gpt-oss-120b62.4%
huggingface.co · 2026-05-15
SWE-bench Pro (Public)Estimated
~#23 of 24 (estimated)~45.9%
- #1Claude Fable 5.293.0%
- #2Claude Opus 5.589.9%
- #22Muse Glimmer51.2%
- ~#23Intern-S2-Preview (35B-A3B)Estimated45.9%
- #23Nex-N2.5-Mini43.8%
No published score — estimated from related benchmarks
DeepSWE v1.1Estimated
~#43 of 46 (estimated)~35.0%
- #1Gemini 4 Argon78.8%
- #2DeepSeek V4.1 Flash 101777.0%
- #42Nex-N2.5-Mini36.1%
- ~#43Intern-S2-Preview (35B-A3B)Estimated35.0%
- #43Hy328.0%
No published score — estimated from related benchmarks
Terminal-Bench 4.0Estimated
~#40 of 102 (estimated)~8.9%
- #1Kimi K3.1 Pro67.2%
- #2Claude Sonnet 5.563.6%
- #39Qwen3.8 2.4T A95B11.1%
- ~#40Intern-S2-Preview (35B-A3B)Estimated8.9%
- #40Muse Spark Open8.1%
No published score — estimated from related benchmarks
LiveCodeBenchEstimated
~#15 of 28 (estimated)~66.0%
- #1Qwen3.8-Omni-Flash92.6%
- #2Qwen3.8 Flash-Next91.9%
- #14Maple-Preview75.1%
- ~#15Intern-S2-Preview (35B-A3B)Estimated66.0%
- #15Mistral Small 464.0%
No published score — estimated from related benchmarks
SciCodeLab-reported
#84 of 11839.6%
- #1Claude Fable 5.267.7%
- #2Claude Opus 5.566.9%
- #82Qwen3.5 122B A10B39.7%
- #84Intern-S2-Preview (35B-A3B)39.6%
- #85gpt-oss-20b38.9%
huggingface.co · 2026-05-15
AA Coding Agent IndexEstimated
~#25 of 26 (estimated)~30.8
- #1Claude Opus 5.561.7
- #2Claude Opus 559.7
- #24GPT-6 Luna31.1
- ~#25Intern-S2-Preview (35B-A3B)Estimated30.8
- #25Muse Spark Open26.3
No published score — estimated from related benchmarks
Writing & chat
IFBenchLab-reported
#7 of 13980.9%
- #1MiniMax-M382.9%
- #2Qwen3.8 2.4T A95B82.8%
- #6Qwen3.8 Flash-Next NVFP481.0%
- #7Intern-S2-Preview (35B-A3B)80.9%
- #8Nemotron Cascade 2 30B A3B80.4%
huggingface.co · 2026-05-15
LMArena Text, Non-English (Elo)Estimated
~#35 of 57 (estimated)~1420 Elo
- #1Gemini 4 ArgonFew votes1512
- #2Claude Opus 5.51497
- #34Gemma 4 26B A4BFew votes1423
- ~#35Intern-S2-Preview (35B-A3B)Estimated1420
- #35Muse GlimmerFew votes1418
No published score — estimated from related benchmarks
LMArena Text (Elo)Estimated
~#35 of 57 (estimated)~1435 Elo
- #1Gemini 4 Argon1525
- #2Claude Opus 5.51507
- #34Qwen3.8 27B1437
- ~#35Intern-S2-Preview (35B-A3B)Estimated1435
- #35MiMo-V2.51434
No published score — estimated from related benchmarks
EQ-Bench Creative Writing v3 (Elo)Estimated
~#13 of 26 (estimated)~1801 Elo
- #1GPT-6 Astra2173
- #2Claude Fable 5.12162
- #12Qwen3.8 2.4T A95B1843
- ~#13Intern-S2-Preview (35B-A3B)Estimated1801
- #13Muse Glimmer1798
No published score — estimated from related benchmarks
Vision
MMMU-ProLab-reported
#29 of 8776.9%
- #1Claude Opus 5.587.7%
- #2GPT-6 Astra86.9%
- #28DeepSeek V4.1 Flash77.0%
- #29Intern-S2-Preview (35B-A3B)76.9%
- #30Grok Build 0.1 061676.5%
huggingface.co · 2026-05-15
CharXiv ReasoningEstimated
~#7 of 12 (estimated)~79.9%
- #1Gemini 3.8 Flash86.2%
- #2Kimi K384.8%
- #6Nemotron 3.5 Super VL80.8%
- ~#7Intern-S2-Preview (35B-A3B)Estimated79.9%
- #7Nemotron 3.5 Super VL NVFP479.9%
No published score — estimated from related benchmarks
OmniDocBench v1.5Estimated
~#4 of 8 (estimated)~86.2%
- #1MiniMax-M391.6%
- #2Kimi K391.1%
- #2Qwen3.8 27B91.1%
- ~#4Intern-S2-Preview (35B-A3B)Estimated86.2%
- #4Gemini 3.1 Pro85.3%
No published score — estimated from related benchmarks
GDP.pdf (AA)Estimated
~#24 of 48 (estimated)~15.5%
- #1Muse Wonder Full32.6%
- #2GPT-6 Astra31.0%
- #23Qwen3.8 Flash-Next15.6%
- ~#24Intern-S2-Preview (35B-A3B)Estimated15.5%
- #24GLM-5.3 Flash15.4%
No published score — estimated from related benchmarks
ChartographyEstimated
~#16 of 25 (estimated)~24.8%
- #1Claude Fable 5.273.4%
- #2GPT-6 Astra71.0%
- #15Gemini 3.1 Pro26.1%
- ~#16Intern-S2-Preview (35B-A3B)Estimated24.8%
- #16Grok 4.620.5%
No published score — estimated from related benchmarks
Hard reasoning
GPQA DiamondEstimated
~#87 of 193 (estimated)~76.3%
- #1GPT-6 Astra96.1%
- #2Gemini 3.8 Flash95.3%
- #86Cogito v2.176.8%
- ~#87Intern-S2-Preview (35B-A3B)Estimated76.3%
- #87Command A+76.1%
No published score — estimated from related benchmarks
Humanity's Last ExamLab-reported
#76 of 19821.9%
- #1Claude Fable 5.274.3%
- #2Claude Opus 5.561.4%
- #74Muse Glimmer22.0%
- #76Intern-S2-Preview (35B-A3B)21.9%
- #77Ring-2.6-1T21.6%
huggingface.co · 2026-05-15
AIME (latest)Estimated
~#15 of 34 (estimated)~88.9%
- #1GLM-5.3 NVFP499.5%
- #2Beam 1.0 501B A23B97.7%
- #14Nova 2.0 Pro Preview89.0%
- ~#15Intern-S2-Preview (35B-A3B)Estimated88.9%
- #15Maple-Preview87.5%
No published score — estimated from related benchmarks
AA Intelligence IndexEstimated
~#52 of 103 (estimated)~24.9
- #1Gemini 4 Carbon59.4
- #2Claude Opus 5.557.6
- #51Inkling25.0
- ~#52Intern-S2-Preview (35B-A3B)Estimated24.9
- #52Ling-3.0-flash-VL24.6
No published score — estimated from related benchmarks