Skip to content
Model Pareto

K2 Horizon 7B

MBZUAI Institute of Foundation ModelsOpen weightsartificialanalysis.ai ↗
Input
—
Output
—
Speed
~110tok/sest.
Context
524Ktokens

Long-tail import from Artificial Analysis (2026-09-24); AA variant: K2 Horizon 7B. 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
BF16 weights
Throughput
~10,073 tok/s
many requests batched
Price
$0.083–0.74 /M tok
busy → light use
On your own machine
RTX 5090 32GB (int4, ~280 tok/s single-stream)
single consumer GPU or Mac
Assumptions (5)
  • BF16 weights (9B params) + 40% KV-cache headroom ≈ 25 GB
  • 1× B200 180GB at $5.98–$14.24/GPU-hour on-demand (2026-09-24)
  • ~10,073 output tok/s aggregate at batch 256 (bandwidth-bound)
  • Blended 3:1 input:output; prefill ~59,063 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

~#23 of 24 (estimated)~46.7%
  1. #1Claude Fable 5.293.0%
  2. #2Claude Opus 5.589.9%
  3. #22Muse Glimmer51.2%
  4. ~#23K2 Horizon 7BEstimated46.7%
  5. #23Nex-N2.5-Mini43.8%
No published score — estimated from related benchmarks
DeepSWE v1.1Estimated
~#43 of 46 (estimated)~34.9%
  1. #1Gemini 4 Argon78.8%
  2. #2DeepSeek V4.1 Flash 101777.0%
  3. #42Nex-N2.5-Mini36.1%
  4. ~#43K2 Horizon 7BEstimated34.9%
  5. #43Hy328.0%
No published score — estimated from related benchmarks
~#25 of 26 (estimated)~28.8
  1. #1Claude Opus 5.561.7
  2. #2Claude Opus 559.7
  3. #24GPT-6 Luna31.1
  4. ~#25K2 Horizon 7BEstimated28.8
  5. #25Muse Spark Open26.3
No published score — estimated from related benchmarks

Hard reasoning

Agents