Llama 4 Scout
- Input
- $0.18/M tok
- Output
- $0.68/M tok
- Blended
- $0.305/M tok · 3:1
- Speed
- 54tok/s
- Context
- 10Mtokens
- Cost per task
- ~$0.29/taskest.
Long-tail import from Artificial Analysis (2026-09-24); AA variant: Llama 4 Scout. Price = AA median host price.
Rankings
Coding
Verified#133 / 137rank by quality15.3±6.5quality / 1003/7 benchmarks measuredWriting & chat
Verified#123 / 162rank by quality24.5±9.5quality / 1001/5 benchmarks measuredVision
Verified#82 / 98rank by quality25.3±8.2quality / 1001/6 benchmarks measuredHard reasoning
Verified#183 / 218rank by quality15.8±3.3quality / 1004/4 benchmarks measuredAgents
Verified#180 / 187rank by quality11.1±5.9quality / 1003/7 benchmarks measured
Self-hostingEstimated
How we estimate →What it would cost to run these open weights yourself on rented GPUs, compared with the API price above.
- Hardware
- 4× H100 80GB
- FP8 weights
- Throughput
- ~9,075 tok/s
- many requests batched
- Price
- $0.14–1.31 /M tok
- busy → light use
- On your own machine
- Mac Studio M5 Max 128GB (int4, ~51 tok/s single-stream)
- single consumer GPU or Mac
Assumptions (5)
- FP8 weights (108.6B params, 17B active per token) + 40% KV-cache headroom ≈ 152 GB
- 4× H100 80GB at $2.69–$6.88/GPU-hour on-demand (2026-09-24)
- ~9,075 output tok/s aggregate at batch 265 (bandwidth-bound); MoE compute scales with active params
- Blended 3:1 input:output; prefill ~110,009 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 VerifiedEstimated
~#19 of 21 (estimated)~59.8%
- #1Claude Opus 596.0%
- #2Claude Sonnet 585.2%
- #18gpt-oss-120b62.4%
- ~#19Llama 4 ScoutEstimated59.8%
- #19NVIDIA Nemotron-Labs Dory1.0 4B59.5%
No published score — estimated from related benchmarks
SWE-bench Pro (Public)Estimated
~#23 of 24 (estimated)~44.8%
- #1Claude Fable 5.293.0%
- #2Claude Opus 5.589.9%
- #22Muse Glimmer51.2%
- ~#23Llama 4 ScoutEstimated44.8%
- #23Nex-N2.5-Mini43.8%
No published score — estimated from related benchmarks
DeepSWE v1.1Estimated
~#43 of 46 (estimated)~31.3%
- #1Gemini 4 Argon78.8%
- #2DeepSeek V4.1 Flash 101777.0%
- #42Nex-N2.5-Mini36.1%
- ~#43Llama 4 ScoutEstimated31.3%
- #43Hy328.0%
No published score — estimated from related benchmarks
Terminal-Bench 4.0Verified
#63 of 1010.0%
- #1Kimi K3.1 Pro67.2%
- #2Claude Sonnet 5.563.6%
- #63Llama 4 Maverick0.0%
- #63Llama 4 Scout0.0%
- #63LongCat 2.00.0%
artificialanalysis.ai · 2026-09-24
LiveCodeBenchVerified
#26 of 2729.9%
- #1Qwen3.8-Omni-Flash92.6%
- #2Qwen3.8 Flash-Next91.9%
- #25Ministral 3 8B30.3%
- #26Llama 4 Scout29.9%
- #27Ministral 3 3B24.7%
artificialanalysis.ai · 2026-09-24
SciCodeVerified
#113 of 11821.3%
- #1Claude Fable 5.267.7%
- #2Claude Opus 5.566.9%
- #112Celeris-121.6%
- #113Llama 4 Scout21.3%
- #114Ministral 3 8B20.7%
artificialanalysis.ai · 2026-09-24
AA Coding Agent IndexEstimated
~#25 of 26 (estimated)~27.5
- #1Claude Opus 5.561.7
- #2Claude Opus 559.7
- #24GPT-6 Luna31.1
- ~#25Llama 4 ScoutEstimated27.5
- #25Muse Spark Open26.3
No published score — estimated from related benchmarks
Writing & chat
IFBenchVerified
#90 of 13939.5%
- #1MiniMax-M382.9%
- #2Qwen3.8 2.4T A95B82.8%
- #89Qwen3 Next 80B A3B Instruct39.7%
- #90Llama 4 Scout39.5%
- #91Olmo 3.1 32B Instruct39.2%
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)~1318 Elo
- #1Gemini 4 ArgonFew votes1512
- #2Claude Opus 5.51497
- #47INTELLECT-3Few votes1327
- ~#48Llama 4 ScoutEstimated1318
- #48Llama 4 Maverick1311
No published score — estimated from related benchmarks
LMArena Text (Elo)Estimated
~#49 of 57 (estimated)~1341 Elo
- #1Gemini 4 Argon1525
- #2Claude Opus 5.51507
- #48Llama 3.1 Nemotron Ultra 253B v1Few votes1348
- ~#49Llama 4 ScoutEstimated1341
- #49Olmo 3.1 32B Instruct1330
No published score — estimated from related benchmarks
EQ-Bench Creative Writing v3 (Elo)Estimated
~#24 of 26 (estimated)~1323 Elo
- #1GPT-6 Astra2173
- #2Claude Fable 5.12162
- #23Gemma 4 31B1368
- ~#24Llama 4 ScoutEstimated1323
- #24gpt-oss-120b961
No published score — estimated from related benchmarks
Vision
MMMU-ProVerified
#70 of 8752.9%
- #1Claude Opus 5.587.7%
- #2GPT-6 Astra86.9%
- #69Nemotron 3 Nano Omni 30B A3B Reasoning53.2%
- #70Llama 4 Scout52.9%
- #70NVIDIA Nemotron Nano 12B v2 VL52.9%
artificialanalysis.ai · 2026-09-24
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)~0.6%
- #1Muse Wonder Full32.6%
- #2GPT-6 Astra31.0%
- #45Ministral 3 8B0.8%
- ~#46Llama 4 ScoutEstimated0.6%
- #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
#139 of 19258.7%
- #1GPT-6 Astra96.1%
- #2Gemini 3.8 Flash95.3%
- #138Olmo 3.1 32B Think59.1%
- #139Llama 4 Scout58.7%
- #140Phi-457.5%
artificialanalysis.ai · 2026-09-24
Humanity's Last ExamVerified
#186 of 1983.8%
- #1Claude Fable 5.274.3%
- #2Claude Opus 5.561.4%
- #186Jamba Reasoning 3B3.8%
- #186Llama 4 Scout3.8%
- #186Phi-43.8%
artificialanalysis.ai · 2026-09-24
Agents
τ²-bench (Telecom)Verified
#92 of 11915.5%
- #1Step 3.7 Flash98.5%
- #2DeepSeek-V4-Pro-0813-nvfp4-DSpark98.3%
- #91Jamba Reasoning 3B15.8%
- #92Llama 4 Scout15.5%
- #93Command A15.2%
artificialanalysis.ai · 2026-09-24
OSWorld-VerifiedEstimated
~#12 of 15 (estimated)~54.3%
- #1Kimi K384.8%
- #2Qwen3.8 27B84.3%
- #11Muse Glimmer65.9%
- ~#12Llama 4 ScoutEstimated54.3%
- #12Nemotron 3.5 Super VL NVFP452.3%
No published score — estimated from related benchmarks
OSWorld 2.0Estimated
~last of 5 (estimated)below measured range (<47.5%)
No published score — estimated from related benchmarks
BrowseCompEstimated
~#20 of 22 (estimated)~45.0%
- #1Atria Dawn Preview92.5%
- #2GPT-6 Astra91.5%
- #19Mistral Medium 3.548.6%
- ~#20Llama 4 ScoutEstimated45.0%
- #20Nemotron 3 Ultra44.4%
No published score — estimated from related benchmarks
GDPval-AA (Elo)Estimated
~#80 of 92 (estimated)~352 Elo
- #1Claude Opus 5.51866
- #2Muse Wonder Full1857
- #79Nova 2.0 Lite354
- ~#80Llama 4 ScoutEstimated352
- #80Mistral Small 4349
No published score — estimated from related benchmarks
AutomationBench-AAVerified
#79 of 840.2%
- #1Claude Fable 5.276.4%
- #2Muse Wonder Full71.3%
- #79Llama 4 Maverick0.2%
- #79Llama 4 Scout0.2%
- #79Magistral Small 1.20.2%
artificialanalysis.ai · 2026-09-24