Kimi Linear 48B A3B Instruct
- Input
- —
- Output
- —
- Speed
- ~100tok/sest.
- Context
- 1Mtokens
Long-tail import from Artificial Analysis (2026-09-24); AA variant: Kimi Linear 48B A3B Instruct. Price = AA median host price.
Rankings
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
- ~7,486 tok/s
- many requests batched
- Price
- $0.051–0.42 /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 (48B params, 3B active per token) + 40% KV-cache headroom ≈ 67 GB
- 1× H200 141GB at $3.59–$7.91/GPU-hour on-demand (2026-09-24)
- ~7,486 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.
Writing & chat
IFBenchVerified
#121 of 13928.1%
- #1MiniMax-M382.9%
- #2Qwen3.8 2.4T A95B82.8%
- #120Hermes 4 - Llama-3.1 70B29.0%
- #121Kimi Linear 48B A3B Instruct28.1%
- #122GLM-4.6V27.9%
artificialanalysis.ai · 2026-09-24
MMMLUEstimated
~last of 5 (estimated)below measured range (<81.3%)
- #1Gemini 3.1 Pro92.6%
- #2Gemma 4 31B88.4%
- #4gpt-oss-120b81.3%
- ~#5Kimi Linear 48B A3B InstructEstimated<81.3%
No published score — estimated from related benchmarks
LMArena Text, Non-English (Elo)Estimated
~#55 of 57 (estimated)~1281 Elo
- #1Gemini 4 ArgonFew votes1512
- #2Claude Opus 5.51497
- #54NVIDIA Nemotron 3 Nano 30B A3B1286
- ~#55Kimi Linear 48B A3B InstructEstimated1281
- #55Phi-41233
No published score — estimated from related benchmarks
LMArena Text (Elo)Estimated
~#55 of 57 (estimated)~1309 Elo
- #1Gemini 4 Argon1525
- #2Claude Opus 5.51507
- #54NVIDIA Nemotron 3 Nano 30B A3B1314
- ~#55Kimi Linear 48B A3B InstructEstimated1309
- #55Olmo 3.1 32B Think1286
No published score — estimated from related benchmarks
EQ-Bench Creative Writing v3 (Elo)Estimated
~#22 of 26 (estimated)~1419 Elo
- #1GPT-6 Astra2173
- #2Claude Fable 5.12162
- #21Gemini 3.1 Pro1491
- ~#22Kimi Linear 48B A3B InstructEstimated1419
- #22Mistral Large 31412
No published score — estimated from related benchmarks
Hard reasoning
GPQA DiamondVerified
#167 of 19241.2%
- #1GPT-6 Astra96.1%
- #2Gemini 3.8 Flash95.3%
- #166Granite 4.0 H Small41.6%
- #167Kimi Linear 48B A3B Instruct41.2%
- #168Gemma 4 E2B40.5%
artificialanalysis.ai · 2026-09-24
Humanity's Last ExamVerified
#198 of 1982.5%
- #1Claude Fable 5.274.3%
- #2Claude Opus 5.561.4%
- #197ERNIE 4.5 300B A47B3.3%
- #198Kimi Linear 48B A3B Instruct2.5%
artificialanalysis.ai · 2026-09-24
AIME (latest)Estimated
~#29 of 34 (estimated)~23.2%
- #1GLM-5.3 NVFP499.5%
- #2Beam 1.0 501B A23B97.7%
- #28Ministral 3 14B30.0%
- ~#29Kimi Linear 48B A3B InstructEstimated23.2%
- #29Ministral 3 3B22.0%
No published score — estimated from related benchmarks
AA Intelligence IndexEstimated
~last of 103 (estimated)below measured range (<3.0)
- #1Gemini 4 Carbon59.4
- #2Claude Opus 5.557.6
- #102K2 Horizon 0.9B3.0
- ~#103Kimi Linear 48B A3B InstructEstimated<3.0
No published score — estimated from related benchmarks
Agents
τ²-bench (Telecom)Verified
#110 of 1190.0%
- #1Step 3.7 Flash98.5%
- #2DeepSeek-V4-Pro-0813-nvfp4-DSpark98.3%
- #110ERNIE 4.5 300B A47B0.0%
- #110Kimi Linear 48B A3B Instruct0.0%
- #110Molmo 7B-D0.0%
artificialanalysis.ai · 2026-09-24
OSWorld-VerifiedEstimated
~#14 of 15 (estimated)~50.8%
- #1Kimi K384.8%
- #2Qwen3.8 27B84.3%
- #13Nemotron 3.5 Super VL51.1%
- ~#14Kimi Linear 48B A3B InstructEstimated50.8%
- #14Claude Haiku 4.550.7%
No published score — estimated from related benchmarks
OSWorld 2.0Estimated
~#3 of 5 (estimated)~53.8%
No published score — estimated from related benchmarks
BrowseCompEstimated
~#21 of 22 (estimated)~32.7%
- #1Atria Dawn Preview92.5%
- #2GPT-6 Astra91.5%
- #20Nemotron 3 Ultra44.4%
- ~#21Kimi Linear 48B A3B InstructEstimated32.7%
- #21Mistral Small 421.3%
No published score — estimated from related benchmarks
GDPval-AA (Elo)Estimated
~#61 of 92 (estimated)~636 Elo
- #1Claude Opus 5.51866
- #2Muse Wonder Full1857
- #60MiniCPM5-2B694
- ~#61Kimi Linear 48B A3B InstructEstimated636
- #61Nemotron 3.5 Lightning624
No published score — estimated from related benchmarks
AutomationBench-AAEstimated
~#79 of 85 (estimated)~0.2%
- #1Claude Fable 5.276.4%
- #2Muse Wonder Full71.3%
- #77Granite 4.2 3B0.4%
- ~#79Kimi Linear 48B A3B InstructEstimated0.2%
- #79gpt-oss-120b0.2%
No published score — estimated from related benchmarks
τ-Bench Banking (AA)Estimated
~last of 91 (estimated)below measured range (<3.3%)
- #1Muse Spark 1.350.5%
- #2GLM-5.350.3%
- #90Llama 4 Scout3.3%
- ~#91Kimi Linear 48B A3B InstructEstimated<3.3%
No published score — estimated from related benchmarks