A.X-K2
- Input
- —
- Output
- —
- Speed
- ~83tok/sest.
- Context
- 262Ktokens
Long-tail import from Artificial Analysis (2026-09-24); AA variant: A.X-K2. 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
- 8× B200 180GB
- FP8 weights
- Throughput
- ~22,549 tok/s
- many requests batched
- Price
- $0.25–2.21 /M tok
- busy → light use
- On your own machine
- Mac Studio M5 Ultra 512GB (int4, ~51 tok/s single-stream)
- single consumer GPU or Mac
Assumptions (5)
- FP8 weights (688B params, 33B active per token) + 25% KV-cache headroom ≈ 860 GB
- 8× B200 180GB at $5.98–$14.24/GPU-hour on-demand (2026-09-24)
- ~22,549 output tok/s aggregate at batch 645 (bandwidth-bound); MoE compute scales with active params
- Blended 3:1 input:output; prefill ~257,727 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
~#14 of 21 (estimated)~71.7%
- #1Claude Opus 596.0%
- #2Claude Sonnet 585.2%
- #13Claude Haiku 4.573.3%
- ~#14A.X-K2Estimated71.7%
- #14Nemotron 3.5 Super VL71.1%
No published score — estimated from related benchmarks
SWE-bench Pro (Public)Estimated
~#23 of 24 (estimated)~50.9%
- #1Claude Fable 5.293.0%
- #2Claude Opus 5.589.9%
- #22Muse Glimmer51.2%
- ~#23A.X-K2Estimated50.9%
- #23Nex-N2.5-Mini43.8%
No published score — estimated from related benchmarks
DeepSWE v1.1Estimated
~#38 of 46 (estimated)~45.3%
- #1Gemini 4 Argon78.8%
- #2DeepSeek V4.1 Flash 101777.0%
- #37Qwen3.8 Max (0902)51.0%
- ~#38A.X-K2Estimated45.3%
- #38Ripple-65B Preview44.3%
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%
- #53Trinity Large Thinking0.5%
- #63A.X-K20.0%
- #63Apodex 1.10.0%
artificialanalysis.ai · 2026-10-07
LiveCodeBenchEstimated
~#15 of 28 (estimated)~64.3%
- #1Qwen3.8-Omni-Flash92.6%
- #2Qwen3.8 Flash-Next91.9%
- #14Maple-Preview75.1%
- ~#15A.X-K2Estimated64.3%
- #15Mistral Small 464.0%
No published score — estimated from related benchmarks
SciCodeVerified
#76 of 11841.0%
- #1Claude Fable 5.267.7%
- #2Claude Opus 5.566.9%
- #75Gemini 3.5 Flash-Lite41.3%
- #76A.X-K241.0%
- #77Trinity Large Thinking40.6%
artificialanalysis.ai · 2026-09-24
AA Coding Agent IndexEstimated
~#25 of 26 (estimated)~30.6
- #1Claude Opus 5.561.7
- #2Claude Opus 559.7
- #24GPT-6 Luna31.1
- ~#25A.X-K2Estimated30.6
- #25Muse Spark Open26.3
No published score — estimated from related benchmarks
Hard reasoning
Humanity's Last ExamVerified
#57 of 19829.6%
- #1Claude Fable 5.274.3%
- #2Claude Opus 5.561.4%
- #56Nemotron 3.5 Super VL30.7%
- #57A.X-K229.6%
- #58Solar Pro 429.2%
artificialanalysis.ai · 2026-09-24
AIME (latest)Estimated
~#8 of 34 (estimated)~93.7%
- #1GLM-5.3 NVFP499.5%
- #2Beam 1.0 501B A23B97.7%
- #7Nova 2.0 Lite94.3%
- ~#8A.X-K2Estimated93.7%
- #8gpt-oss-120b93.4%
No published score — estimated from related benchmarks
Agents
τ²-bench (Telecom)Estimated
~#44 of 120 (estimated)~66.7%
- #1Step 3.7 Flash98.5%
- #2DeepSeek-V4-Pro-0813-nvfp4-DSpark98.3%
- #43Nemotron 3 Super 120B A12B67.8%
- ~#44A.X-K2Estimated66.7%
- #44gpt-oss-120b65.8%
No published score — estimated from related benchmarks
OSWorld-VerifiedEstimated
~#12 of 15 (estimated)~64.6%
- #1Kimi K384.8%
- #2Qwen3.8 27B84.3%
- #11Muse Glimmer65.9%
- ~#12A.X-K2Estimated64.6%
- #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
~#19 of 22 (estimated)~63.3%
- #1Atria Dawn Preview92.5%
- #2GPT-6 Astra91.5%
- #18Nemotron 3.5 Bolt 74B-A9B70.0%
- ~#19A.X-K2Estimated63.3%
- #19Mistral Medium 3.548.6%
No published score — estimated from related benchmarks
GDPval-AA (Elo)Estimated
~#31 of 92 (estimated)~1268 Elo
No published score — estimated from related benchmarks
AutomationBench-AAEstimated
~#45 of 85 (estimated)~17.0%
- #1Claude Fable 5.276.4%
- #2Muse Wonder Full71.3%
- #44Qwen3.7 Plus17.4%
- ~#45A.X-K2Estimated17.0%
- #45Ling-3.0-flash-VL15.7%
No published score — estimated from related benchmarks