Skip to content
Model Pareto

K2-V2

MBZUAI Institute of Foundation ModelsOpen weightsartificialanalysis.ai ↗
Input
—
Output
—
Speed
~85tok/sest.
Context
512Ktokens

Long-tail import from Artificial Analysis (2026-09-24); AA variant: K2-V2 (high). 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× B200 180GB
FP8 weights
Throughput
~2,343 tok/s
many requests batched
Price
$0.35–3.09 /M tok
busy → light use
On your own machine
Mac Studio M5 Max 128GB (int4, ~12 tok/s single-stream)
single consumer GPU or Mac
Assumptions (5)
  • FP8 weights (70B params) + 40% KV-cache headroom ≈ 98 GB
  • 1× B200 180GB at $5.98–$14.24/GPU-hour on-demand (2026-09-24)
  • ~2,343 output tok/s aggregate at batch 67 (bandwidth-bound)
  • Blended 3:1 input:output; prefill ~15,188 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

Hard reasoning

Agents

~#12 of 15 (estimated)~55.7%
  1. #1Kimi K384.8%
  2. #2Qwen3.8 27B84.3%
  3. #11Muse Glimmer65.9%
  4. ~#12K2-V2Estimated55.7%
  5. #12Nemotron 3.5 Super VL NVFP452.3%
No published score — estimated from related benchmarks
OSWorld 2.0Estimated
~#3 of 5 (estimated)~53.8%
  1. #1GLM-5.3 Flash59.1%
  2. #2Kimi K358.3%
  3. ~#3K2-V2Estimated53.8%
  4. #3GPT-5.6 Terra50.2%
No published score — estimated from related benchmarks
BrowseCompEstimated
~#21 of 22 (estimated)~41.9%
  1. #1Atria Dawn Preview92.5%
  2. #2GPT-6 Astra91.5%
  3. #20Nemotron 3 Ultra44.4%
  4. ~#21K2-V2Estimated41.9%
  5. #21Mistral Small 421.3%
No published score — estimated from related benchmarks
~#55 of 92 (estimated)~757 Elo
  1. #1Claude Opus 5.51866
  2. #2Muse Wonder Full1857
  3. #54Gemini 3.1 Pro776
  4. ~#55K2-V2Estimated757
  5. #55Qwen3.7 Plus756
No published score — estimated from related benchmarks
~#47 of 85 (estimated)~7.5%
  1. #1Claude Fable 5.276.4%
  2. #2Muse Wonder Full71.3%
  3. #46K2 Horizon 3.7B14.7%
  4. ~#47K2-V2Estimated7.5%
  5. #47Qwen3.5 397B A17B6.9%
No published score — estimated from related benchmarks
~#61 of 91 (estimated)~11.1%
  1. #1Muse Spark 1.350.5%
  2. #2GLM-5.350.3%
  3. #60K-EXAONE 2.0 080311.5%
  4. ~#61K2-V2Estimated11.1%
  5. #61Devstral Small 210.7%
No published score — estimated from related benchmarks